# 10. - AI Features # AI Overview & Licensing ## Study Notes | Topic | Description | |---|---| | Genesys AI | Suite of artificial intelligence capabilities integrated across PureCloud | | Purpose | Enhance agent productivity, automate interactions, improve customer experience | | Licensing Model | AI features available as add-on modules to Premium edition | | Deployment | Cloud-native, no infrastructure required | | Scope | Covers analytics, automation, agent assist, and autonomous agents | --- ## Navigation Admin → Organization Settings → Licensing → AI Modules OR Admin → Billing & Subscriptions → AI Products --- ## Genesys AI Suite Overview Genesys PureCloud AI is an integrated suite of artificial intelligence capabilities designed to enhance every aspect of contact center operations. From agent assistance to autonomous automation, AI is embedded throughout the platform. ### Core AI Components - **Customer Insights** - Advanced analytics and AI-powered quality management - **Agent Copilot** - Real-time intelligent agent assistance - **Predictive Routing** - AI-optimized contact-to-agent matching - **Virtual Agent Flows** - Autonomous conversational AI agents - **Workforce Optimization** - AI-driven forecasting and scheduling - **Interaction Analytics** - Deep conversation analysis and insights - **Quality Management** - AI-assisted quality evaluation ### Key Benefits - **Increased Productivity** - Agents handle more contacts with guidance - **Better Decisions** - Data-driven insights inform strategy - **Cost Reduction** - Automation reduces operational expenses - **Improved Quality** - AI ensures consistency and compliance - **Enhanced Experience** - Faster resolution and personalized service - **Risk Mitigation** - Proactive compliance and fraud detection --- ## AI Licensing Model ### Premium Edition Requirement ``` Base: Premium Edition (Required) ↓ + AI Add-on Modules (Optional) ├── Customer Insights Module ├── Workforce Optimization Module ├── Virtual Agent Module └── Contact Center Intelligence Module ↓ = Complete AI-Powered Solution ``` ### Licensing Structure | Component | Type | Cost Model | Required | |---|---|---|---| | Premium Edition | Base license | Per-user monthly | Yes | | Customer Insights | Add-on module | Per-user or pool | No (recommended) | | Workforce Optimization | Add-on module | Organizational | No | | Virtual Agent | Add-on module | Per-agent license | No | | Advanced Analytics | Add-on module | Organizational | No | --- ## Study Notes - AI Capabilities by Module | Module | Primary Capability | Sub-features | License Type | |---|---|---|---| | Customer Insights | Interaction analytics and quality management | Sentiment analysis, topic detection, quality evaluation, agent coaching | Per-user | | Workforce Optimization | Forecasting, scheduling, and performance management | Demand forecasting, schedule optimization, workforce analytics, productivity tracking | Organizational | | Agent Copilot | Real-time agent assistance | Knowledge recommendations, script guidance, sentiment alerts, next-action suggestions | Per-user | | Virtual Agent | Autonomous conversation handling | Intent recognition, multi-turn dialogue, escalation logic, omnichannel support | Per-agent | | Predictive Routing | Intelligent contact routing | Skill matching, availability prediction, performance prediction, load balancing | Included in WFO | | Contact Center Intelligence | Deep conversation insights | Speech analytics, text analytics, compliance monitoring, competitive intelligence | Organizational | | Advanced Analytics | Custom reporting and dashboards | Business analytics, predictive metrics, custom visualizations | Organizational | --- ## AI Capabilities by Use Case ### For Agents ``` Agent Experience Enhancements: ├── Agent Copilot │ ├── Real-time knowledge suggestions │ ├── Script guidance │ ├── Sentiment monitoring │ └── Next-action recommendations ├── Predictive Routing │ ├── Optimal contact matching │ ├── Skill-based distribution │ └── Workload balancing └── Performance Insights ├── Real-time coaching ├── Quality feedback └── Training recommendations ``` ### For Supervisors ``` Supervisor Experience Enhancements: ├── Workforce Optimization │ ├── Real-time dashboards │ ├── Agent performance metrics │ ├── Coaching opportunities │ └── Compliance alerts ├── Customer Insights │ ├── Quality management │ ├── Interaction analytics │ ├── Team performance trends │ └── Automated evaluations └── Advanced Analytics ├── Custom reports ├── Predictive metrics ├── Trend analysis └── Business intelligence ``` ### For Customers ``` Customer Experience Enhancements: ├── Faster Resolution │ ├── Agent Copilot speeds answers │ ├── Virtual agents 24/7 availability │ └── Predictive routing reduces wait ├── Better Experience │ ├── Skilled agent matching │ ├── Personalized service │ ├── Appropriate escalation │ └── Consistent quality └── Self-Service Options ├── Virtual agent automation ├── Knowledge base access ├── Portal access └── Omnichannel support ``` ### For Managers/Executives ``` Management Experience Enhancements: ├── Business Intelligence │ ├── Advanced analytics dashboards │ ├── Custom reporting │ ├── Predictive forecasting │ └── Trend analysis ├── Strategic Planning │ ├── Capacity forecasting │ ├── ROI measurement │ ├── Competitive intelligence │ └── Market insights └── Performance Optimization ├── Cost analysis ├── Efficiency metrics ├── Quality improvement └── Customer satisfaction tracking ``` --- ## Implementation Guide ### Step 1: Assessment & Planning 1. Audit current contact center operations 2. Identify pain points and improvement opportunities 3. Document baseline metrics (AHT, FCR, CSAT, costs) 4. Assess readiness for AI adoption 5. Define success metrics and KPIs 6. Estimate ROI for each module 7. Plan implementation timeline ### Step 2: Module Selection 1. Evaluate all available AI modules 2. Prioritize based on business needs 3. Assess agent and supervisor readiness 4. Determine implementation sequence 5. Calculate total cost of ownership 6. Secure budget approval 7. Obtain licenses from Genesys ### Step 3: Foundation Setup 1. Ensure Premium Edition active 2. Purchase selected AI modules 3. Assign licenses to users 4. Set up knowledge management system 5. Configure backend integrations 6. Enable audit logging and monitoring 7. Establish governance policies ### Step 4: Module-Specific Configuration 1. **Customer Insights**: Configure quality evaluation rules 2. **Workforce Optimization**: Set up forecasting models 3. **Agent Copilot**: Populate knowledge base 4. **Virtual Agent**: Design conversation flows 5. **Predictive Routing**: Define skills and proficiency levels 6. **Advanced Analytics**: Create custom dashboards 7. **Contact Center Intelligence**: Enable analytics engines ### Step 5: Training & Change Management 1. Develop training curriculum 2. Train supervisors first 3. Train agents on their tools 4. Provide ongoing support 5. Gather feedback early 6. Address concerns and resistance 7. Celebrate early wins ### Step 6: Phased Rollout 1. Start with single department/queue 2. Monitor metrics closely 3. Gather user feedback 4. Optimize based on learnings 5. Expand to additional queues 6. Scale across entire organization 7. Continuous improvement process --- ## How to Implement | Phase | Description | Timeline | Modules | |---|---|---|---| | Planning | Assess needs, select modules, plan approach | Week 1-2 | All | | Foundation | Set up licenses, integrations, governance | Week 2-4 | All | | Configuration | Configure each module's settings | Week 4-6 | Module-specific | | Training | Educate teams on features and benefits | Week 6-7 | All | | Pilot | Deploy to test group, monitor metrics | Week 7-8 | All | | Rollout | Expand to production, scale up | Week 8-10 | All | | Optimization | Monitor, tune, improve continuously | Week 10+ | All | --- ## Genesys AI Platform Architecture ``` Genesys PureCloud AI Platform ┌─────────────────────────────────────────────────┐ │ Core PureCloud Platform │ │ (Voice, Chat, Email, Social, Video) │ └──────────────┬──────────────────────────────────┘ │ ┌───────┴────────┐ │ │ ┌──────▼──────┐ ┌──────▼──────────┐ │ Interaction │ │ Agent/Customer │ │ Processing │ │ Data │ └──────┬──────┘ └────────┬────────┘ │ │ └──────────┬───────┘ │ ┌─────────▼──────────────┐ │ AI/ML Engines │ │ │ ├─ NLP & Intent Engine │ ├─ Sentiment Analysis │ ├─ Topic Detection │ ├─ Entity Recognition │ ├─ Predictive Models │ └─ Recommendation Engine │ │ ┌────┴─────────────────────────────────────┐ │ AI Module Applications │ │ │ ├─ Customer Insights │ │ ├─ Interaction Analytics │ │ ├─ Quality Evaluation │ │ ├─ Compliance Monitoring │ │ └─ Coaching Recommendations │ │ │ ├─ Agent Copilot │ │ ├─ Knowledge Recommendations │ │ ├─ Script Guidance │ │ ├─ Sentiment Alerts │ │ └─ Next Action Suggestions │ │ │ ├─ Predictive Routing │ │ ├─ Skill Matching │ │ ├─ Load Balancing │ │ ├─ Performance Prediction │ │ └─ Availability Analysis │ │ │ ├─ Virtual Agent Flows │ │ ├─ Conversation Management │ │ ├─ Intent Recognition │ │ ├─ Escalation Logic │ │ └─ Multi-turn Dialogue │ │ │ ├─ Workforce Optimization │ │ ├─ Forecasting │ │ ├─ Scheduling │ │ ├─ Performance Analytics │ │ └─ Productivity Tracking │ │ │ ├─ Contact Center Intelligence │ │ ├─ Speech Analytics │ │ ├─ Text Analytics │ │ ├─ Conversation Mining │ │ └─ Competitive Intelligence │ │ │ └─ Advanced Analytics │ ├─ Custom Dashboards │ ├─ Predictive Metrics │ ├─ Business Intelligence │ └─ Data Visualization │ │ └─────────────────────────────────────────────┘ ``` --- ## AI Modules Detailed Comparison ### Customer Insights Module ``` Primary Use: Quality Management & Interaction Analytics Features: ├── Interaction Recording & Analysis ├── Sentiment Detection (Real-time & historical) ├── Topic Detection (Automatic issue categorization) ├── Speech Analytics (Word/phrase analysis) ├── Quality Evaluation (AI-assisted scoring) ├── Compliance Monitoring (Regulation checking) ├── Agent Coaching (Targeted improvements) └── Custom Metrics (Business-specific analysis) Pricing: Per-user monthly ($XX-XXX depending on tier) ROI: 10-20% quality improvement, reduced audit burden Best For: Quality-focused, compliance-heavy organizations Typical Users: QA teams, supervisors, compliance officers ``` ### Workforce Optimization Module ``` Primary Use: Forecasting, Scheduling, Performance Analytics Features: ├── Demand Forecasting (Historical + predictive) ├── Schedule Optimization (Auto-scheduling) ├── Workforce Analytics (Performance metrics) ├── Productivity Tracking (Real-time dashboards) ├── Absence & Leave Management ├── Skills-based Workforce Planning ├── Performance Management └── Adherence Monitoring Pricing: Organizational license (negotiated) ROI: 5-15% labor cost reduction, improved efficiency Best For: Large organizations, complex scheduling needs Typical Users: Workforce planners, managers, analysts ``` ### Agent Copilot Module ``` Primary Use: Real-Time Agent Assistance Features: ├── Knowledge Recommendations ├── Script Guidance ├── Sentiment Monitoring ├── Next Action Suggestions ├── Real-time Coaching ├── Performance Insights ├── Learning Reinforcement └── Omnichannel Support Pricing: Per-user monthly ($XX-XXX depending on tier) ROI: 10-30% AHT reduction, 15-25% FCR improvement Best For: Agent productivity, customer satisfaction Typical Users: All agents, supervisors ``` ### Virtual Agent Module ``` Primary Use: Autonomous Customer Interaction Handling Features: ├── Conversational AI ├── Intent Recognition ├── Multi-turn Dialogue ├── Transaction Processing ├── Escalation Logic ├── 24/7 Availability ├── Omnichannel Support └── Sentiment Awareness Pricing: Per-virtual agent (negotiated) ROI: 60-80% cost reduction for automated interactions Best For: High-volume routine interactions Typical Users: Customers, supervisors (monitoring) ``` ### Advanced Analytics Module ``` Primary Use: Business Intelligence & Custom Reporting Features: ├── Custom Dashboard Creation ├── Predictive Analytics ├── Trend Analysis ├── Business Intelligence ├── Data Visualization ├── Scheduled Reports ├── Data Export └── Integration APIs Pricing: Organizational license (negotiated) ROI: Better decision-making, strategic insights Best For: Data-driven organizations, large enterprises Typical Users: Managers, executives, analysts ``` --- ## AI Licensing Structure Comparison ### Small Organization (50 agents) ``` Premium Edition: 50 users x $XX/month = $XXXX Agent Copilot: 50 users x $XX/month = $XXXX Customer Insights: 10 supervisors x $XX/month = $XXXX Workforce Optimization: Organizational = $XXXX Total Monthly: $XXXX Cost per Agent: $XXX/month ROI: 12-18 months through efficiency gains ``` ### Mid-Market (200 agents) ``` Premium Edition: 200 users x $XX/month = $XXXX Agent Copilot: 200 users x $XX/month = $XXXX Customer Insights: 50 supervisors x $XX/month = $XXXX Workforce Optimization: Organizational = $XXXX Virtual Agent: 5 agents x $XXXX/month = $XXXX Advanced Analytics: Organizational = $XXXX Total Monthly: $XXXX Cost per Agent: $XXX/month ROI: 6-12 months through automation + efficiency ``` ### Enterprise (1000+ agents) ``` Premium Edition: 1000 users x $XX/month = $XXXXX Agent Copilot: 1000 users x $XX/month = $XXXXX Customer Insights: 200 supervisors x $XX/month = $XXXXX Workforce Optimization: Organizational = $XXXXX Virtual Agent: 20 agents x $XXXX/month = $XXXXX Contact Center Intelligence: Organizational = $XXXXX Advanced Analytics: Organizational = $XXXXX Total Monthly: $XXXXX Cost per Agent: $XXX/month ROI: 3-6 months through massive automation + efficiency ``` --- ## Implementation Roadmap ### Phase 1: Foundation (Month 1-2) ``` Quick Wins: ├── Enable Predictive Routing │ └─ Immediate improvement in routing efficiency ├── Enable Customer Insights │ └─ Gain visibility into interaction quality └── Activate Agent Copilot └─ First-touch improvement in agent effectiveness Expected Impact: ├── Routing efficiency: +10-15% ├── Agent confidence: +20-30% ├── FCR improvement: +5-10% └── Cost per contact: -5-10% ``` ### Phase 2: Intelligence (Month 2-4) ``` Expansion: ├── Deploy Workforce Optimization │ └─ Optimize scheduling and forecasting ├── Enhance Customer Insights │ └─ Add compliance monitoring └── Optimize Agent Copilot └─ Improve knowledge base quality Expected Impact: ├── Labor efficiency: +10-15% ├── Quality improvement: +15-20% ├── FCR improvement: +10-20% └── CSAT improvement: +10-15% ``` ### Phase 3: Automation (Month 4-6) ``` Advanced: ├── Deploy Virtual Agent Flows │ └─ Automate routine interactions ├── Activate Advanced Analytics │ └─ Enable strategic decision-making └── Integrate all modules └─ Seamless intelligence platform Expected Impact: ├── Automation rate: 50-70% of routine volume ├── Cost reduction: 30-50% ├── Customer satisfaction: +15-25% └── Operational efficiency: +25-35% ``` --- ## Real-World Implementation Timeline ### Week 1-2: Assessment & Planning ``` Day 1-3: ├─ Kick-off meeting ├─ Current state assessment ├─ Identify pain points └─ Document baseline metrics Day 4-10: ├─ Module evaluation ├─ ROI analysis ├─ Risk assessment ├─ Develop implementation plan Day 11-14: ├─ Secure approvals ├─ License procurement ├─ Team assignments └─ Detailed project plan ``` ### Week 3-4: Foundation Setup ``` Day 15-21: ├─ Activate Premium Edition ├─ Purchase AI modules ├─ License assignment ├─ Infrastructure setup Day 22-28: ├─ Knowledge base creation ├─ Integration testing ├─ Security validation └─ Documentation ``` ### Week 5-6: Configuration ``` Day 29-35: ├─ Customer Insights setup ├─ Quality rules configuration ├─ Analytics dashboards └─ Coaching workflows Day 36-42: ├─ Agent Copilot setup ├─ Knowledge population ├─ Sentiment analysis └─ Relevance tuning ``` ### Week 7-8: Training & Pilot ``` Day 43-49: ├─ Supervisor training ├─ Agent training ├─ Pilot queue selection └─ Monitoring setup Day 50-56: ├─ Pilot deployment ├─ Daily monitoring ├─ Feedback collection ├─ Quick optimizations └─ Success validation ``` ### Week 9-10: Rollout ``` Day 57-63: ├─ Production deployment ├─ Agent onboarding ├─ Supervisor monitoring ├─ Support availability Day 64-70: ├─ Scale to all queues ├─ Monitor close for issues ├─ Gather feedback └─ Celebrate successes ``` --- ## AI Features by Job Role ### For Agents | Feature | Module | Benefit | |---|---|---| | Real-time knowledge suggestions | Agent Copilot | Faster resolution, fewer transfers | | Script guidance | Agent Copilot | Better quality, compliance adherence | | Sentiment alerts | Agent Copilot | De-escalation, higher satisfaction | | Optimal contact matching | Predictive Routing | Better skill match, faster resolution | | Performance insights | Customer Insights | Learning and improvement | | Interaction recording | Customer Insights | Quality and coaching | ### For Supervisors | Feature | Module | Benefit | |---|---|---| | Real-time agent dashboards | Workforce Optimization | Team visibility and coaching | | Quality evaluation | Customer Insights | Objective performance assessment | | Automated coaching | Customer Insights | Targeted development | | Schedule optimization | Workforce Optimization | Better coverage, agent satisfaction | | Compliance monitoring | Customer Insights | Risk mitigation | | Performance trends | Advanced Analytics | Identify patterns and opportunities | ### For Managers | Feature | Module | Benefit | |---|---|---| | Demand forecasting | Workforce Optimization | Budget planning, capacity management | | Cost analysis | Workforce Optimization | ROI tracking, budget optimization | | Team performance analytics | Customer Insights | Performance visibility | | Custom dashboards | Advanced Analytics | Strategic decision-making | | Predictive insights | Advanced Analytics | Proactive management | | Business intelligence | Advanced Analytics | Competitive advantage | ### For Executives | Feature | Module | Benefit | |---|---|---| | Revenue impact analysis | Advanced Analytics | Business value justification | | Cost savings tracking | Workforce Optimization | ROI demonstration | | Customer satisfaction trends | Customer Insights | Service quality visibility | | Market intelligence | Contact Center Intelligence | Competitive positioning | | Strategic dashboards | Advanced Analytics | Executive visibility | | Predictive metrics | Advanced Analytics | Forward-looking planning | --- ## Best Practices for AI Implementation ### Governance & Management - **Clear ownership** - Assign AI program owner and team - **Success metrics** - Define measurable KPIs before implementation - **Change management** - Plan for organizational change - **Training program** - Invest in comprehensive user education - **Governance policies** - Establish AI usage guidelines - **Regular reviews** - Monthly assessment and optimization ### Technical Implementation - **Phased approach** - Start with one module, expand gradually - **Integration planning** - Ensure backend system connectivity - **Data quality** - Ensure clean, accurate data for AI models - **Security protocols** - Implement proper access controls - **Audit trails** - Enable logging for compliance and learning - **Monitoring** - Set up dashboards for tracking performance ### User Adoption - **Executive sponsorship** - Strong leadership support - **Early adopters** - Start with enthusiastic teams - **Success stories** - Share wins to build confidence - **Continuous learning** - Regular training and tips - **Feedback loops** - Listen to user concerns - **Support availability** - Easy access to help and troubleshooting ### Continuous Improvement - **Daily monitoring** - Track key metrics continuously - **Weekly reviews** - Assess performance and issues - **Monthly optimization** - Tune settings and rules - **Quarterly planning** - Assess and plan enhancements - **Annual strategy** - Review overall AI strategy and ROI - **Learning loop** - Capture insights to improve AI models --- ## Common Challenges & Solutions | Challenge | Solution | |---|---| | Agents skeptical of AI | Demonstrate time-saving benefits, celebrate wins | | Knowledge base quality | Establish content review process, regular updates | | Integration complexity | Partner with Genesys professional services | | High implementation cost | Show ROI, phased approach to spread costs | | Change resistance | Strong change management, executive support | | Slow adoption | User-friendly interfaces, comprehensive training | | AI accuracy issues | Improve training data, tune models | | Compliance concerns | Implement proper audit trails and controls | | Technical challenges | Dedicated technical support and monitoring | | Low usage of features | Training, communication, usage incentives | --- ## Licensing & Compliance ### License Tracking ``` System automatically tracks: ├─ Active AI module users ├─ Virtual agent instances in use ├─ Monthly consumption metrics ├─ Grace period usage └─ Compliance status Monthly Report includes: ├─ License utilization rate ├─ Cost per user/module ├─ Usage trends ├─ Recommendations for optimization └─ Billing details ``` ### Compliance Monitoring - **Audit logs** - All AI actions tracked and logged - **Data protection** - Customer data encrypted and protected - **Regulatory compliance** - GDPR, CCPA, HIPAA support - **Access controls** - Role-based permission system - **Interaction recording** - Captured for compliance - **Report generation** - Automated compliance reporting --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What is Genesys AI? | Suite of AI capabilities across PureCloud (Copilot, Routing, Virtual Agent, etc.) | | What edition is required? | Premium Edition (AI modules are add-ons) | | What modules are available? | Customer Insights, Workforce Optimization, Agent Copilot, Virtual Agent, Advanced Analytics | | How is AI licensed? | Per-user or organizational depending on module | | What's the expected ROI? | 6-18 months depending on implementation scope | | Which module should we start with? | Predictive Routing or Agent Copilot for quick impact | | Can we implement gradually? | Yes, phased approach recommended | | What's most important for success? | Quality change management and user training | | How do we measure AI impact? | Track FCR, AHT, CSAT, cost per contact, automation rate | | What about compliance concerns? | Audit trails, access controls, and monitoring built-in | | Can we customize AI models? | Yes, tuning available; may require Genesys services for deep customization | | What's the implementation timeline? | 8-12 weeks for full deployment, quick wins in 2-4 weeks | | How do we ensure adoption? | Executive support, training, early wins, continuous communication | | What's the cost per agent per month? | $XX-XXX depending on modules and scale | | Where do we start? | Assessment → Licensing → Foundation → Phased Implementation | --- ## Key Takeaways - **Integrated Suite** - Genesys AI is not standalone products but integrated capabilities - **Premium Requirement** - Premium Edition is required; AI modules are add-ons - **Flexible Licensing** - Mix and match modules based on needs - **Phased Implementation** - Start with high-impact modules, expand gradually - **Significant ROI** - Typical payback of 6-18 months - **Change Critical** - Success depends more on change management than technology - **Omnichannel Ready** - AI works across all communication channels - **Continuous Learning** - AI models improve with more data and interaction - **Compliance Built-In** - Security, audit, and compliance features included - **Quick Wins Possible** - Can see improvements within 2-4 weeks of implementation --- ## Getting Started Checklist ### Pre-Implementation - [ ] Conduct current state assessment - [ ] Document baseline metrics - [ ] Identify pain points and opportunities - [ ] Evaluate all AI modules - [ ] Calculate ROI for each module - [ ] Secure budget and approvals - [ ] Assemble implementation team - [ ] Plan change management approach ### Implementation Preparation - [ ] Purchase Premium Edition - [ ] License required AI modules - [ ] Assign licenses to users - [ ] Set up knowledge management system - [ ] Plan integrations - [ ] Establish governance policies - [ ] Create training curriculum - [ ] Set up monitoring dashboards ### Deployment - [ ] Deploy to pilot group - [ ] Monitor closely - [ ] Gather feedback - [ ] Optimize configuration - [ ] Scale to production - [ ] Provide ongoing support - [ ] Track metrics - [ ] Plan continuous improvement --- ## Additional Resources ### Official Documentation Links - Genesys Cloud AI Overview: https://help.genesys.com/genesyscloud/current/en-us/AI_Overview.html - AI Module Licensing: https://help.genesys.com/genesyscloud/current/en-us/AI_Licensing.html - Architect Documentation: https://help.genesys.com/genesyscloud/current/en-us/Architect.html - Customer Insights Guide: https://help.genesys.com/genesyscloud/current/en-us/CustomerInsights.html ### Support Contacts - Genesys Sales: sales@genesys.com - Genesys Support: https://support.genesys.com - AI Services Team: ai-services@genesys.com - Community Forums: https://community.genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation **Version:** 1.0 # AI Studio & AI Guides # Genesys PureCloud AI Studio & AI Guides Documentation ## Study Notes | Topic | Description | |---|---| | AI Studio | Centralized workbench for building, managing, and deploying AI-powered experiences | | AI Guides | No-code feature that converts business instructions into Virtual Agents using natural language | | Purpose | Enable business users to create intelligent virtual agents without coding expertise | | Licensing | Requires Genesys Cloud AI Experience tokens (metered pricing) | | Innovation Level | Level 4 agentic AI - semi-autonomous with defined boundaries and guardrails | --- ## Navigation Admin → AI Studio OR Admin → Contact Center → Automation → AI Studio --- ## AI Studio Overview Genesys Cloud AI Studio provides a centralized workbench to build, manage and deploy AI experiences like Virtual Agents and Agent Copilots. It serves as the command center for organizations to create next-generation AI-powered customer experiences with governance, control, and scalability. ### Key Capabilities - Natural language editor with no-code and low-code tools for users without technical know-how - Unified environment for designing AI agents across all channels - Built-in governance and compliance controls - AI Guides that allow business teams to create Virtual Agents that respond intelligently to customer context and adapt their behavior dynamically within conversations - Customizable summaries for interactions - Integration with existing Architect Virtual Agent flows - Version control and deployment management - Performance analytics and monitoring ### Strategic Positioning AI Studio represents a leap from simple execution to intelligent problem-solving - Level 4 where semi-autonomous systems are configured around specific objectives using reasoning, planning and memory to figure out how best to accomplish goals while still operating within clearly defined boundaries to meet compliance and policy requirements --- ## Edition & Module Requirements | Requirement | Details | |---|---| | Minimum Edition | Genesys Cloud CX 1, CX 2, CX 3, or CX 4 license | | Licensing Model | AI Experience tokens (metered, usage-based pricing) | | Permissions | Role-based AI Studio permissions required | | Additional | Access to Virtual Agent module for full functionality | --- ## Study Notes - AI Studio Components | Component | Purpose | User Type | |---|---|---| | AI Guides | Create Virtual Agents from natural language prompts | Business users, CX teams | | Guide Editor | Build and refine guide instructions and logic | All user levels | | Virtual Agent Integration | Deploy guides to Virtual Agent flows | Technical users | | Customizable Summaries | Shape interaction summaries to business needs | Admins, supervisors | | Governance Controls | Ensure compliance and policy adherence | Admins, compliance | | Performance Dashboard | Monitor guide usage and effectiveness | Managers, analysts | | API Access | Programmatic guide and agent management | Developers | --- ## AI Guides Overview Genesys Cloud AI Guides enables business users to create AI-powered Virtual Agents through natural language instructions without coding, facilitating conversational flows that mirror customer journeys, adapt dynamically to customer context, and combine structured logic with AI capabilities ### How AI Guides Work 1. Business user describes goal in plain language or uploads process documentation 2. AI Guides use large language models (LLMs)-based natural language processing to interpret user prompts and documents, generating complete agentic flows including intents, slots and dialog logic 3. System generates draft virtual agent with conversation flow 4. User reviews, edits, and customizes the generated guide 5. AI executable instructions produced by AI Guides are fully editable within Genesys Cloud, allowing users to update messages, logic and backend integrations before publishing 6. Guide is published and connected to Virtual Agent 7. Virtual Agent uses guide to handle customer interactions ### Key Features - Natural Language, No Code Required - easily build or refine virtual agents using plain language or existing documentation with no coding skills needed - Build Once, Deploy Anywhere - design experiences once and deploy them across Genesys Cloud Virtual Agent and Copilots, and more to maintain consistency and reduce duplication of effort - Enterprise-Grade Collaboration - seamlessly connect front, middle and back-office systems to execute tasks, automate workflows and deliver measurable business outcomes - Guardrails Built In - implement configurable, testable safety controls to help enable increased accuracy, appropriate tone and policy compliance to support responsible adoption of agentic AI - Knowledge integration - guides can use either knowledge Workbench v2 knowledge bases or knowledge fabric configurations to answer customer questions at any point in a conversation --- ## Implementation Guide ### Step 1: Assessment & Planning 1. Identify suitable use cases for AI Guides 2. Document business processes and customer journeys 3. Gather process documentation or playbooks 4. Define success metrics for each guide 5. Plan escalation scenarios 6. Assess team readiness for AI automation 7. Plan change management approach ### Step 2: Licensing & Setup 1. Ensure organization has Genesys Cloud CX license (CX 1, CX 2, CX 3, or CX 4) 2. Purchase Genesys Cloud AI Experience tokens 3. Add necessary AI Studio permissions 4. Set up role-based access controls 5. Configure Virtual Agent if not already enabled 6. Test integration with backend systems 7. Establish governance policies ### Step 3: Guide Creation 1. Navigate to AI Studio 2. Create a guide using an AI prompt, convert a process document into a guide, or create from scratch by starting with a blank guide 3. Describe goal in natural language or upload documentation 4. Review AI-generated guide structure 5. Edit guide instructions and customize flow 6. Add variables and data integrations 7. Configure escalation paths ### Step 4: Testing & Refinement 1. Preview guide behavior in test environment 2. Author preview before publish - preview real knowledge responses during Guide configuration to confirm accuracy and behavior 3. Test with sample customer scenarios 4. Verify escalation triggers work correctly 5. Validate data integrations 6. Test across all supported channels 7. Gather feedback from SMEs ### Step 5: Publishing & Deployment 1. Publish guide to Virtual Agent 2. Connect the guide to Virtual Agent flows in Architect 3. Assign to production queues 4. Monitor initial interactions closely 5. Validate customer experience 6. Adjust guide parameters based on feedback 7. Scale to additional queues as needed ### Step 6: Monitoring & Optimization 1. Monitor guide usage metrics daily 2. Track customer satisfaction and resolution rates 3. Review escalation patterns 4. Analyze customer feedback 5. Refine guide instructions based on data 6. A/B test different guide variations 7. Update regularly based on learnings --- ## How to Implement | Phase | Description | Timeline | |---|---|---| | Planning | Identify use cases, document processes, assess readiness | Week 1-2 | | Setup | Activate licenses, configure permissions, enable integrations | Week 2-3 | | Creation | Build guides from prompts or documents, test | Week 3-5 | | Testing | Validate behavior, test scenarios, refine | Week 5-6 | | Pilot | Deploy to select queues, monitor closely | Week 6-7 | | Rollout | Expand to production, scale across organization | Week 7-8 | | Optimization | Monitor, analyze, improve continuously | Ongoing | --- ## AI Studio & AI Guides Architecture ``` AI Studio Centralized Workbench ┌─────────────────────────────────────────┐ │ AI Studio Command Center │ ├─────────────────────────────────────────┤ │ │ │ Guide Creation Interface │ │ ├─ Natural Language Prompt Input │ │ ├─ Document Upload & Conversion │ │ └─ Blank Guide Builder │ │ │ │ AI Generation Engine │ │ ├─ LLM Processing │ │ ├─ Intent Detection │ │ ├─ Slot Identification │ │ └─ Dialog Logic Generation │ │ │ │ Guide Editor & Customization │ │ ├─ Instruction Editing │ │ ├─ Variable Management │ │ ├─ Logic Configuration │ │ └─ Integration Setup │ │ │ │ Governance & Controls │ │ ├─ Built-in Guardrails │ │ ├─ Compliance Checking │ │ ├─ Tone Configuration │ │ └─ Brand Alignment Rules │ │ │ │ Publishing & Deployment │ │ ├─ Version Control │ │ ├─ Publish to Virtual Agent │ │ ├─ Queue Assignment │ │ └─ Channel Distribution │ │ │ │ Analytics & Monitoring │ │ ├─ Usage Dashboards │ │ ├─ Performance Metrics │ │ ├─ Escalation Tracking │ │ └─ Feedback Collection │ │ │ │ System Integration │ │ ├─ Architect Virtual Agent │ │ ├─ Knowledge Management │ │ ├─ Backend Data Systems │ │ └─ Customer Data Platforms │ │ │ └─────────────────────────────────────────┘ ↓ Virtual Agent Deployment ↓ Customer Interactions ``` --- ## AI Guides Use Cases & Examples ### Use Case 1: Order Status & Tracking ``` Business Process Documentation: 1. Collect order number from customer 2. Look up order in system 3. Provide tracking information 4. Offer additional help options 5. Close or escalate as needed AI Guide Generated: Utterances: ├─ "Where's my order?" ├─ "Track my package" ├─ "Order status" └─ "When will my order arrive?" Intents: ├─ Track Order ├─ Delivery Status └─ Tracking Number Dialog Flow: ├─ Ask for order number ├─ Query order database ├─ Provide shipping status ├─ Offer further assistance └─ Escalate if needed Resolution Rate: 85-90% (self-service) ``` ### Use Case 2: Password Reset Process ``` Uploaded Process Document: "Follow these steps for password reset: 1. Verify customer identity with security questions 2. Confirm email address on file 3. Send password reset link 4. Confirm reset completed 5. Offer additional support" AI Guide Generated: Intents: ├─ Password Reset Request ├─ Account Access Issue └─ Security Verification Dialog: ├─ "I'll help you reset your password" ├─ "Let me verify who you are" ├─ Customer answers security questions ├─ System confirms identity ├─ "Check your email for reset link" ├─ "Did you successfully reset?" └─ Provide follow-up support Resolution Rate: 92-95% (self-service) ``` ### Use Case 3: Appointment Scheduling ``` Prompt Input: "Create a guide for customers to book service appointments. Must: - Ask for service type - Show available dates/times - Confirm appointment - Send confirmation" AI Guide Generated: Intents: ├─ Schedule Appointment ├─ Change Appointment └─ Cancel Appointment Dialog: ├─ "What service do you need?" ├─ "When works best for you?" ├─ Show available slots ├─ "Let me confirm..." ├─ Send calendar confirmation ├─ Offer reschedule option Resolution Rate: 88-92% (self-service) ``` ### Use Case 4: Billing Question & Payment ``` Process Flow: "Customers with billing questions: 1. Verify account 2. Explain charges 3. Offer payment options 4. Process if customer agrees 5. Send receipt/confirmation" AI Guide Generated: Intents: ├─ Check Bill Amount ├─ Explain Charges ├─ Make Payment └─ Dispute Charge Dialog: ├─ Confirm customer identity ├─ "Let me look up your account" ├─ Display current balance ├─ "Would you like to pay?" ├─ Secure payment processing ├─ Send receipt and confirmation ├─ Escalate disputes Resolution Rate: 80-85% (self-service) ``` --- ## Modular Guides Within Virtual Agents AI Guides are designed to create modular bot flows that can be combined within a single virtual agent - each guide typically addresses a specific use case like order tracking, password reset or appointment booking, and these flows can be added to your virtual agent's overall configuration ### Multi-Guide Architecture Example ``` Virtual Agent: Customer Service Bot ├─ Guide 1: Order Tracking │ ├─ Intents: Track order, delivery status │ ├─ Resolution: Order lookup + tracking info │ └─ Escalation: Complex shipment issues │ ├─ Guide 2: Account Management │ ├─ Intents: Password reset, update info │ ├─ Resolution: Self-service account changes │ └─ Escalation: Security concerns │ ├─ Guide 3: Billing & Payments │ ├─ Intents: Check bill, make payment │ ├─ Resolution: Payment processing │ └─ Escalation: Billing disputes │ └─ Guide 4: Appointment Booking ├─ Intents: Schedule, reschedule, cancel ├─ Resolution: Appointment management └─ Escalation: Complex scheduling needs Router (AI determines which guide needed): Customer: "Where's my order?" → Route to Guide 1: Order Tracking → Resolve: Self-service tracking info Customer: "I want to schedule a service" → Route to Guide 4: Appointment Booking → Resolve: Appointment confirmation Customer: "I have a billing question" → Route to Guide 3: Billing & Payments → Resolve or escalate accordingly ``` --- ## Knowledge Integration with AI Guides In a future release, Genesys Cloud will allow AI Guides to use knowledge articles to answer customer questions while continuing through a defined process - guides can answer customer questions using approved knowledge content at any point in a conversation and then continue the task without losing context ### Knowledge-Aware Guide Benefits ``` Before Knowledge Integration: Customer: "What's your return policy?" Guide: "Let me connect you with an agent who can answer policy questions" → Escalation (unnecessary) After Knowledge Integration: Customer: "What's your return policy?" Guide: [Accesses knowledge base] "Here's our return policy... Now, back to your order tracking, I see your item was delivered..." → Continues conversation naturally → No escalation needed ``` ### Configuration Options Guide authors can choose to inherit knowledge from the connected Virtual Agent or select a Guide-specific Knowledge source Flexible knowledge sourcing: - Use Virtual Agent's knowledge base - Configure guide-specific knowledge source - Support for knowledge Workbench v2 - Support for knowledge fabric configurations --- ## Real Flow Scenario: AI Guide in Action ``` Customer Interaction Example: Customer Calls: "I need to reset my password" ↓ Virtual Agent Answers (AI Guide: Account Management) "Hi! I'm here to help. To reset your password, I'll need to verify your identity. What's your email address on file?" ↓ Customer: "john.smith@email.com" ↓ Guide Verifies: Customer email matches account ↓ Guide: "Thanks. What was your first pet's name?" ↓ Customer: "Fluffy" ↓ Guide Confirms: Security answer matches ↓ Guide: "Perfect! I've sent a password reset link to your email. Please check your inbox and click the link to set a new password." ↓ Customer: "Got it, thanks!" ↓ Guide: "You're welcome. Is there anything else I can help with today?" ↓ Customer: "No, that's all" ↓ Guide: "Have a great day!" ↓ Interaction Complete: ├─ Resolution: Self-service password reset ├─ Tokens Used: 1-2 per interaction ├─ Customer Satisfaction: High (immediate help) └─ Cost: ~$0.05-0.15 per interaction ``` --- ## Licensing & Token Pricing ### AI Experience Token Model Organizations need Genesys Cloud AI Experience tokens to use AI Studio, with tokens based pricing model to monitor feature usage and consumption ### Token Consumption AI Guides consume tokens for each interaction session and for each guide created ### Pricing Considerations - **Per-interaction tokens** - Consumed when customers interact with guides - **Per-guide tokens** - Consumed when creating new guides - **Metered model** - Pay for what you use - **Flexibility** - Tokens can scale up or down based on demand - **Multiple LLM support** - AI Studio is compatible with proprietary, open source and Amazon Bedrock large language models (LLMs) and advanced frontier models from companies such as OpenAI, Anthropic and Google ### Token Consumption Example ``` Small Organization: ├─ 3 AI Guides created (3 guides × tokens) ├─ 500 customer interactions/month (500 × tokens) ├─ Monthly token usage: ~750 tokens └─ Estimated cost: $XX-XXX/month Mid-Market Organization: ├─ 10 AI Guides created (10 guides × tokens) ├─ 5,000 customer interactions/month (5,000 × tokens) ├─ Monthly token usage: ~5,100 tokens └─ Estimated cost: $XXX-XXXX/month Enterprise Organization: ├─ 50 AI Guides created (50 guides × tokens) ├─ 50,000 customer interactions/month (50,000 × tokens) ├─ Monthly token usage: ~50,100 tokens └─ Estimated cost: $XXXX-XXXXX/month ``` --- ## Best Practices ### Guide Design - **Clear intent** - Define specific goal for each guide - **Natural language** - Write instructions as you would for an employee - **Process documentation** - Use existing playbooks and procedures - **Modular approach** - Create focused guides for specific tasks - **Escalation paths** - Define clear handoff scenarios - **Testing** - Always test with real scenarios before production - **Iteration** - Guides improve with updates and refinement ### Guardrails & Governance - Implement configurable, testable safety controls to help enable increased accuracy, appropriate tone and policy compliance to support responsible adoption of agentic AI - Set boundaries for agent autonomy - Define tone and brand voice - Establish compliance requirements - Monitor sensitive operations - Implement audit trails - Regular compliance reviews ### Knowledge Management - Author preview before publish - preview real knowledge responses during Guide configuration to confirm accuracy and behavior - Keep knowledge articles current - Ensure accurate information - Update for policy changes - Test knowledge accuracy - Monitor article usage - Gather feedback from guides using knowledge ### Performance Optimization - Monitor token usage closely - Track guide effectiveness metrics - Analyze escalation reasons - Gather customer feedback - Refine guides based on data - A/B test different approaches - Regular guide audits and updates --- ## Common Implementation Scenarios ### Scenario 1: Quick Start (Small Team) ``` Timeline: 2-3 weeks Setup: ├─ 2-3 simple guides ├─ Basic integration (order lookup, FAQ) ├─ Limited escalation paths └─ Single channel deployment Expected Results: ├─ 60-70% automation of routine inquiries ├─ Quick implementation, minimal training ├─ Fast ROI (4-6 weeks) └─ Monthly token usage: ~300-500 ``` ### Scenario 2: Comprehensive (Mid-Market) ``` Timeline: 6-8 weeks Setup: ├─ 8-12 specialized guides ├─ Multiple backend integrations ├─ Complex escalation logic ├─ Omnichannel deployment └─ Knowledge base integration Expected Results: ├─ 50-65% automation of customer volume ├─ Significant cost savings ├─ 8-12 week ROI └─ Monthly token usage: ~3,000-5,000 ``` ### Scenario 3: Enterprise (Full Scale) ``` Timeline: 12-16 weeks Setup: ├─ 30-50+ specialized guides ├─ Full system integration ├─ Advanced routing and logic ├─ Multi-language support ├─ Comprehensive knowledge integration └─ Global channel deployment Expected Results: ├─ 40-60% automation of global volume ├─ Major cost reduction ├─ Continuous optimization ├─ 6-10 week ROI └─ Monthly token usage: ~20,000-50,000+ ``` --- ## Troubleshooting Guide | Issue | Cause | Resolution | |---|---|---| | Guide doesn't understand customer intent | Insufficient training variations | Add more example utterances to training data | | Escalations happening too frequently | Over-strict guardrails or incomplete logic | Expand guide capabilities and relax thresholds | | Inaccurate information from knowledge | Stale knowledge articles | Update knowledge base and preview before publish | | Slow response time | Token consumption or system latency | Optimize guide logic and integrations | | Integration failures | Backend system connectivity issues | Verify API connections and data endpoints | | Guides not deployed to queues | Missing Virtual Agent configuration | Check Architect flow configuration and deployment | | Customer dissatisfaction | Poor conversation quality | Refine guide instructions and tone | | Token overages | Guides consuming more than expected | Analyze usage patterns and optimize interactions | | Editing feels clunky | Unfamiliar with new guide model | Review updated guide model documentation | | Knowledge not accessible in guides | Knowledge not properly configured | Configure knowledge source and test access | --- ## AI Guides vs. Traditional Bot Flows | Feature | AI Guides | Traditional Flows | |---|---|---| | Creation Method | Natural language prompts | Manual design | | Skill Required | Business knowledge | Technical/coding | | Speed to Deploy | Days to weeks | Weeks to months | | Iteration Time | Minutes to hours | Hours to days | | Intent Recognition | AI-powered, flexible | Rule-based, rigid | | Conversation Quality | Natural, contextual | Scripted, menu-driven | | Maintenance | AI learns from data | Manual updates | | Integration | Automatic via AI | Manual configuration | | Cost to Build | Low (no developers) | High (developer time) | | Scaling | Easy, rapid deployment | Time-consuming | --- ## Real-World Timeline Example ### Week 1-2: Planning & Assessment ``` Day 1-3: Kickoff and planning ├─ Identify 3-5 use cases ├─ Gather process documentation └─ Assess team readiness Day 4-10: Documentation review ├─ Refine process flows ├─ Identify integration needs ├─ Plan escalation scenarios Day 11-14: Setup and licensing ├─ Purchase AI Experience tokens ├─ Configure permissions └─ Set up integrations ``` ### Week 3-4: Guide Creation ``` Day 15-20: First guide creation ├─ Upload process documentation ├─ Review AI-generated guide ├─ Customize and refine └─ Connect to Virtual Agent Day 21-28: Additional guides ├─ Create guides 2-4 ├─ Refine based on feedback ├─ Configure escalation logic └─ Integration testing ``` ### Week 5-6: Testing & Refinement ``` Day 29-35: Comprehensive testing ├─ Test all guide scenarios ├─ Validate integrations ├─ Test escalation paths └─ Gather feedback Day 36-42: Optimization ├─ Refine guide instructions ├─ Adjust confidence thresholds ├─ Optimize knowledge integration └─ Performance tuning ``` ### Week 7-8: Deployment ``` Day 43-49: Pilot deployment ├─ Deploy to pilot queue ├─ Monitor closely ├─ Gather customer feedback └─ Make adjustments Day 50-56: Full rollout ├─ Expand to all queues ├─ Monitor metrics ├─ Provide agent support └─ Celebrate success ``` --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What is AI Studio? | Centralized workbench for building, managing, and deploying AI experiences | | What is an AI Guide? | No-code feature that converts business instructions into Virtual Agents using AI | | What licensing is required? | Genesys Cloud CX license + AI Experience tokens | | How do AI Guides work? | Users describe goals in plain language, AI generates guide structure, user customizes | | Can non-technical users create guides? | Yes, AI Guides are designed for business users without coding skills | | What level of agentic AI is this? | Level 4 - semi-autonomous with defined boundaries and guardrails | | How long to create a guide? | Minutes to hours depending on complexity | | Can guides be combined? | Yes, multiple guides can be modular components in one Virtual Agent | | How are guides customized? | Fully editable - users can change messages, logic, and integrations | | What about knowledge integration? | Guides can access knowledge articles to answer questions within conversation | | What's the pricing model? | Token-based - consumption tracked per interaction and per guide created | | Can guides handle escalation? | Yes, with configurable escalation paths to human agents | | What channels do guides support? | All Genesys channels - voice, chat, email, messaging, etc. | | How do guides improve over time? | AI learns from interactions; user refines guides based on metrics | | What's the expected ROI timeline? | 4-12 weeks depending on implementation scope | --- ## Key Takeaways - **No-Code Creation** - Natural language, no code required - easily build or refine virtual agents using plain language or existing documentation with no coding skills needed - **Intelligent Automation** - AI-powered guides intelligently handle customer conversations with reasoning and planning - **Enterprise Control** - Guardrails built in - implement configurable, testable safety controls to help enable increased accuracy, appropriate tone and policy compliance - **Rapid Deployment** - Create and deploy guides in days, not months - **Modular Design** - Build once, deploy anywhere across virtual agents, copilots and more to maintain consistency and reduce duplication of effort - **Token-Based Pricing** - Pay only for what you use with flexible scaling - **Knowledge Integration** - Guides can answer customer questions using approved knowledge content during any step of a process - **Business User Friendly** - Designed for CX teams, not IT/developers - **Continuous Learning** - Guides improve as they handle more interactions - **Compliance Ready** - Built-in governance and audit capabilities --- ## Customizable Summaries (Additional AI Studio Feature) Customizable Summaries allows admins to shape interaction summaries to fit their business needs, enhancing consistency, compliance, and operation efficiency across human and AI agent interactions ### Use Cases - Tailor summaries to specific business requirements - Ensure compliance with regulatory standards - Align summaries with operational priorities - Support multiple languages and regions - Customize for different teams and use cases --- ## Getting Started Checklist ### Pre-Implementation - [ ] Assess current contact center operations - [ ] Identify 3-5 suitable use cases - [ ] Gather process documentation and playbooks - [ ] Determine success metrics for each guide - [ ] Assess team readiness for AI automation - [ ] Plan change management approach ### Licensing & Setup - [ ] Purchase Genesys Cloud AI Experience tokens - [ ] Assign AI Studio permissions to team - [ ] Configure role-based access controls - [ ] Enable Virtual Agent module - [ ] Set up backend system integrations - [ ] Test knowledge base connectivity ### Guide Development - [ ] Create first AI Guide from process document - [ ] Test guide functionality thoroughly - [ ] Customize instructions and logic - [ ] Configure escalation paths - [ ] Connect to Virtual Agent - [ ] Deploy to pilot queue ### Monitoring & Optimization - [ ] Monitor guide performance daily - [ ] Track customer satisfaction metrics - [ ] Analyze escalation patterns - [ ] Refine guides based on data - [ ] Scale to additional queues - [ ] Plan continuous improvements --- ## Additional Resources ### Official Documentation Links - AI Studio Overview: help.genesys.cloud/articles/about-ai-studio/ - AI Guides Overview: help.genesys.cloud/articles/ai-guides-overview/ - Knowledge Integration: help.genesys.cloud/announcements/knowledge-integration-for-ai-guides/ - AI Studio Permissions: help.genesys.cloud/articles/ai-studio-permissions/ ### Support Contacts - Genesys Sales: sales@genesys.com - Genesys Support: https://support.genesys.com - AI Services Team: ai-services@genesys.com - Community Forums: https://community.genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation (help.genesys.cloud) **Validated:** Current with January-March 2026 releases **Version:** 1.0 # Agentic Virtual Agents # Genesys PureCloud Agentic Virtual Agents Documentation ## Study Notes | Topic | Description | |---|---| | Agentic Virtual Agents | AI-powered autonomous agents capable of reasoning, planning, and taking action | | Core Technology | Generative AI with natural language understanding and decision-making | | Autonomy Level | Semi-autonomous with configurable boundaries and safety guardrails | | Channels | Voice, chat, email, messaging, social media | | Capability | Handle complex multi-step customer interactions independently | --- ## Navigation Admin → Contact Center → Virtual Agents OR Admin → AI Studio → Virtual Agents --- ## Agentic Virtual Agents Overview Agentic Virtual Agents represent the next evolution of conversational AI - autonomous agents capable of reasoning between actions and knowledge to solve customer problems intelligently within defined boundaries. Unlike traditional bots that follow rigid decision trees, agentic agents use generative AI to understand context, make decisions, and dynamically adapt their behavior. ### Key Capabilities - **Intelligent Reasoning** - AI thinks through problems and determines best solutions - **Autonomous Action** - Takes actions independently without scripted flows - **Context Understanding** - Maintains conversation context across multiple turns - **Knowledge Integration** - Accesses and applies knowledge dynamically - **System Integration** - Executes transactions and backend actions - **Learning Capability** - Improves responses based on interaction outcomes - **Safety Guardrails** - Operates within defined policies and compliance boundaries - **Omnichannel Support** - Works seamlessly across all communication channels ### How They Differ from Traditional Bots ``` Traditional Bot: ├─ Rigid decision trees ├─ Menu-driven interactions ├─ Limited context awareness ├─ Fixed scripted responses ├─ Capability boundaries unclear └─ Must escalate at first complexity Agentic Virtual Agent: ├─ Intelligent reasoning and planning ├─ Natural conversational flow ├─ Full context preservation ├─ Dynamic adaptive responses ├─ Clear defined boundaries └─ Handles complex scenarios autonomously ``` --- ## Edition & Module Requirements | Requirement | Details | |---|---| | Minimum Edition | Genesys Cloud CX 1, CX 2, CX 3, or CX 4 license | | Module | Virtual Agent module with AI capabilities | | Licensing | AI Experience tokens (metered usage-based) | | AI Studio | Access to AI Studio for guide creation and management | | Knowledge | Knowledge management system for agent reference | --- ## Study Notes - Agentic Virtual Agent Components | Component | Purpose | Function | |---|---|---| | Natural Language Engine | Understands customer intent | Processes speech/text input | | Reasoning Engine | Determines best solution approach | Evaluates options and decides action | | Knowledge Base | Reference material for answers | Provides accurate information | | Action Executor | Performs required actions | Executes system tasks and transactions | | Context Manager | Maintains conversation history | Preserves information across turns | | Guardrail Controller | Enforces boundaries | Prevents unauthorized actions | | Learning Module | Improves over time | Analyzes outcomes for optimization | | Integration Layer | Connects to backend systems | Access to CRM, billing, ticketing, etc. | --- ## Agentic Virtual Agent Architecture ``` Customer Contact ↓ Agentic Virtual Agent Entry ├── Voice (IVR entry point) ├── Chat Interface ├── Email System ├── Messaging Platform └── Social Media ↓ Natural Language Understanding ├── Speech-to-Text (voice) ├── Intent Recognition (what customer wants) ├── Entity Extraction (who, what, when, where) ├── Sentiment Analysis └── Context Comprehension ↓ Reasoning & Planning Engine ├── Evaluate customer request ├── Assess available capabilities ├── Consider guardrails and constraints ├── Determine optimal approach ├── Plan multi-step solution └── Prepare action sequence ↓ Knowledge Access Layer ├── Search knowledge base ├── Retrieve relevant information ├── Evaluate accuracy and applicability ├── Prepare response content └── Maintain knowledge context ↓ Decision Point: Can Handle? ├── YES: Execute independently │ ├── Perform required actions │ ├── Fetch real-time data │ ├── Execute transactions │ ├── Generate response │ └── Provide solution │ └── NO: Escalate to Human ├── Complex/ambiguous scenario ├── Policy exception needed ├── High-risk action ├── Customer preference └── Warm handoff with full context ↓ Response Generation ├── Natural Language Generation ├── Tone and Brand Voice ├── Clarity and Conciseness └── Accessibility Compliance ↓ Conversation Continuation ├── Verify customer satisfaction ├── Offer additional assistance ├── Provide relevant next steps ├── Document interaction └── Learn from outcome ``` --- ## Agentic Capabilities & Functions ### Autonomous Decision Making ``` Customer: "I've been trying to cancel my subscription for weeks but the website won't let me" Agent Analysis: ├─ Understand: Customer frustration, technical issue ├─ Assess: Can perform cancellation autonomously ├─ Evaluate: Business rules for cancellation ├─ Decide: Proceed with cancellation + retention offer └─ Plan: Execute cancellation, offer alternative plan Agent Response: "I understand your frustration. I can cancel your subscription right now. Before I do, I see you've been with us for 3 years. Can I offer you a discounted plan as an alternative? What would help?" ``` ### Multi-Step Problem Solving ``` Customer: "I want to change my billing address and also apply the promotion I saw in email" Agent Reasoning: ├─ Task 1: Update billing address ├─ Task 2: Apply promotional code ├─ Task 3: Verify changes in system ├─ Task 4: Confirm new total with customer └─ Task 5: Send updated confirmation Agent Execution: "I can help with both of those. Let me update your address, apply the promotion, and show you the impact on your bill: Old address: 123 Main St New address: 456 Oak Ave Current plan: $X/month With promotion: $Y/month Monthly savings: $Z Should I go ahead with these changes?" ``` ### Intelligent Escalation ``` Customer: "I need a refund for faulty merchandise" Agent Assessment: ├─ Evaluate: Refund policy ├─ Check: Purchase history ├─ Assess: Complexity (straightforward refund) ├─ Decide: Can handle independently └─ Execute: Process refund Response: "I can definitely help with that. I see your purchase on Feb 10. You're within our 30-day return window. I'm issuing a full refund now. You'll see the credit in 3-5 business days. Would you like to arrange a return shipment?" --- vs. --- Customer: "I need a refund but it's been 45 days and there's a special circumstance..." Agent Assessment: ├─ Evaluate: Outside standard 30-day policy ├─ Assess: Requires exception approval ├─ Determine: Needs human judgment ├─ Decide: Escalate to supervisor └─ Execute: Warm handoff Response: "I understand - you have a special circumstance that's outside our standard 30-day window. Let me connect you with a supervisor who can review options. One moment..." ``` ### Dynamic Adaptation ``` Interaction 1 - Standard Path: Agent: "I can help with your order issue. What's your order number?" Customer: "12345" Agent: [Looks up order, provides status] Interaction 2 - Emotional/Frustrated: Agent: [Detects frustration in tone] Agent: "I hear this has been frustrating. I'm personally going to help you resolve this right now..." Interaction 3 - Knowledgeable Customer: Agent: [Detects technical language] Agent: "I can see you're familiar with our system. Here are the technical details of your issue..." ``` --- ## Real Flow Scenarios: Agentic Virtual Agents in Action ### Scenario 1: Complex Order Issue Resolution ``` Customer Calls: "I ordered item X last week but received item Y instead" Timeline: 10:00:05 AM - Call Connected to Agentic Agent ↓ 10:00:10 - Agent Processes Input ├─ Understands: Wrong item received ├─ Identifies: Potential shipping error └─ Assesses: Solvable by agent ↓ 10:00:15 - Agent Takes Action ├─ Query: Order database ├─ Retrieve: Order 12345 details ├─ Analysis: Item Y in stock, Item X available ├─ Evaluate: Best resolution for customer └─ Decision: Replace + expedited shipping ↓ 10:00:45 - Agent Communicates "I found the issue - you received item Y instead of item X. I'm shipping item X to you via expedited delivery at no charge. You should receive it in 2 days. You can keep item Y as our apology. Good?" Customer: "Yes, that works!" ↓ 10:00:55 - Agent Executes ├─ Initiate replacement shipment ├─ Process return label for wrong item ├─ Apply goodwill credit ├─ Send confirmation email └─ Document resolution ↓ 10:01:00 AM - Issue Resolved Resolution: Self-service (no escalation) Time to Resolution: 55 seconds Cost: ~$5-10 in shipping + goodwill Customer Satisfaction: High (proactive solution) ``` ### Scenario 2: Billing Dispute with Research ``` Customer: "I was charged twice for my subscription" Agent Process: 10:00 AM - Analysis ├─ Retrieve account ├─ Review transactions ├─ Identify: Two charges on same date ├─ Assess: Billing system error (likely) └─ Reason: Needs investigation but obvious 10:05 AM - Investigation ├─ Check system logs ├─ Verify: Duplicate transaction confirmed ├─ Research: Processing glitch identified ├─ Escalate?: No - clear error, can resolve └─ Decide: Refund one charge immediately 10:10 AM - Resolution "I found the issue - our system charged you twice due to a processing error. I'm immediately refunding the duplicate charge of $99.99. You'll see it within 2-3 business days. I'm also applying a $20 service credit for the inconvenience. Is there anything else I can help?" Customer: "That's great, thank you!" Result: ├─ Issue resolved autonomously ├─ No escalation needed ├─ Customer satisfied ├─ Cost: $99.99 refund + $20 credit = $119.99 └─ Outcome: Retained customer, quick resolution ``` ### Scenario 3: Proactive Problem Prevention ``` Situation: New customer during implementation phase Agent Monitoring: ├─ Tracks: Customer account setup ├─ Identifies: Configuration incomplete ├─ Predicts: Customer may hit issues └─ Acts: Reaches out proactively Agent Initiates: "Hi Sarah, I noticed your account setup is almost complete, but I see you haven't configured your team members yet. Would you like help doing that now? I can walk you through it in 2 minutes." Customer: "Sure, that would help!" Agent Guides: ├─ Explains: Setup process ├─ Configures: Team members ├─ Tests: Access ├─ Confirms: All working └─ Provides: Next steps doc Result: ├─ Prevented future support tickets ├─ Improved onboarding experience ├─ Increased early adoption └─ Reduced support costs ``` ### Scenario 4: Multi-Intent Conversation ``` Customer: "Hi, I need to update my payment method, schedule a service appointment, and ask about your loyalty program" Agent Processing: ├─ Identifies: 3 separate intents ├─ Prioritizes: Payment method > appointment > info ├─ Plans: Multi-step interaction sequence └─ Executes: Handles all in single conversation Flow: Agent: "I can help with all of those. Let me start with updating your payment method." [Updates payment method] Agent: "Perfect. Now let's schedule your service appointment. What dates work?" [Books appointment] Agent: "Great! Quick question - you asked about our loyalty program. You actually qualify for Gold tier based on your spending!" [Explains benefits, enrolls automatically] Result: ├─ All 3 issues resolved ├─ Time: Single 5-minute call ├─ Customer: One conversation, multiple solutions └─ Efficiency: What would take 3 transfers now takes 1 ``` --- ## Agentic Virtual Agent Capabilities Matrix ### By Interaction Complexity ``` Simple Tasks (Agent Handles 95%+): ├─ Account status inquiries ├─ FAQ and knowledge searches ├─ Password resets ├─ Basic transactions ├─ Simple scheduling └─ Resolution Rate: 92-97% Medium Complexity (Agent Handles 70-85%): ├─ Billing adjustments ├─ Order modifications ├─ Appointment changes ├─ Return processing ├─ Plan upgrades └─ Resolution Rate: 70-85% Complex Tasks (Agent Handles 40-60%): ├─ Billing disputes ├─ Complaints/resolution ├─ Custom solutions ├─ Policy exceptions ├─ Escalations needed └─ Resolution Rate: 40-60% ``` ### By Business Function ``` Customer Service: ├─ Order tracking ├─ Returns & replacements ├─ Troubleshooting ├─ Status inquiries └─ Resolution: 80-90% Billing/Payments: ├─ Payment processing ├─ Invoice inquiries ├─ Billing corrections ├─ Payment plans └─ Resolution: 75-85% Appointments/Scheduling: ├─ Schedule appointment ├─ Reschedule/cancel ├─ Find availability ├─ Send reminders └─ Resolution: 85-95% Sales/Retention: ├─ Upsell opportunities ├─ Plan recommendations ├─ Offer promotions ├─ Prevent churn └─ Resolution: 70-80% Technical Support: ├─ Basic troubleshooting ├─ Access issues ├─ Configuration help ├─ Advanced escalation └─ Resolution: 65-75% ``` --- ## Agentic Guardrails & Safety Controls ### Built-in Safety Mechanisms ``` Guardrail Levels: CRITICAL (Always Block): ├─ Data breach attempts ├─ Fraud patterns ├─ Unauthorized access ├─ Security violations └─ Action: Immediate block + escalate HIGH (Require Approval): ├─ Large financial transactions ├─ Account terminations ├─ Policy exceptions ├─ Sensitive data access └─ Action: Require verification MEDIUM (Monitor Closely): ├─ Refunds over threshold ├─ Plan downgrades ├─ Churn risk actions ├─ Unusual patterns └─ Action: Execute with logging LOW (Standard Operation): ├─ Routine transactions ├─ Information requests ├─ Schedule changes ├─ Status updates └─ Action: Execute normally ``` ### Configurable Boundaries ``` Organization Can Define: ├─ Maximum transaction amount (agent handle) ├─ Which actions require approval ├─ Policy exception rules ├─ Escalation triggers ├─ Communication tone standards ├─ Retention/discount limits ├─ Data access restrictions └─ Compliance requirements ``` --- ## Agentic Virtual Agent vs. Traditional Virtual Agent | Feature | Agentic Agent | Traditional Agent | |---|---|---| | Decision Making | Reasoning-based | Rule-based | | Conversation Flow | Dynamic, adaptive | Scripted, fixed | | Context Understanding | Full multi-turn | Limited | | Problem Solving | Independent reasoning | Pre-defined paths | | Complexity Handling | Handles moderate complexity | Limited scenarios | | Adaptation | Real-time behavior changes | No adaptation | | Learning | Improves from outcomes | Static | | Escalation Judgment | Intelligent assessment | Pre-set triggers | | Customer Experience | Natural, conversational | Menu-driven | | Autonomy | Semi-autonomous | Fully scripted | | Setup Complexity | Moderate (AI handles) | Low | | Flexibility | High | Low | | Time to Deploy | Days-weeks | Weeks-months | | Resolution Rate | 50-70% complex issues | 30-50% complex | --- ## Implementation Roadmap ### Phase 1: Foundation (Weeks 1-2) ``` Activities: ├── Audit current virtual agent capabilities ├── Identify top 3-5 use cases for agentic ├── Assess customer journey complexity ├── Define guardrails and boundaries ├── Plan integration with existing systems └── Gather requirements from stakeholders Deliverables: ├── Use case prioritization matrix ├── Guardrail policy document ├── Integration requirements list └── Success metric definitions ``` ### Phase 2: Development (Weeks 3-5) ``` Activities: ├── Build/migrate use cases to agentic agents ├── Configure AI Guides for customer flows ├── Set up knowledge base integration ├── Configure guardrails and safety controls ├── Integrate backend systems └── Set up monitoring and analytics Deliverables: ├── Agentic agents built and configured ├── Integration tests completed ├── Guardrail validation completed ├── Monitoring dashboards created └── Documentation complete ``` ### Phase 3: Testing & Optimization (Weeks 6-7) ``` Activities: ├── Comprehensive scenario testing ├── Load and performance testing ├── Guardrail effectiveness testing ├── Integration validation ├── Customer experience testing └── Security and compliance validation Deliverables: ├── Test results and sign-off ├── Performance baselines ├── Optimization recommendations ├── Final readiness confirmation └── Issue resolution log ``` ### Phase 4: Pilot Deployment (Week 8) ``` Activities: ├── Deploy to pilot queue/segment ├── Monitor interactions closely (24/7) ├── Gather customer feedback ├── Track key performance metrics ├── Make rapid optimizations └── Prepare for full rollout Deliverables: ├── Daily performance reports ├── Customer feedback summary ├── Optimization log ├── Pilot success metrics └── Full rollout plan ``` ### Phase 5: Full Rollout (Weeks 9-10) ``` Activities: ├── Expand to remaining queues ├── Monitor close for issues ├── Provide agent training (monitor mode) ├── Support team readiness ├── Scale gradually based on confidence └── Establish optimization cadence Deliverables: ├── Rollout completion checklist ├── Production metrics ├── Team training completion ├── Ongoing monitoring setup └── Optimization plan ``` ### Phase 6: Optimization & Learning (Ongoing) ``` Activities: ├── Daily metric monitoring ├── Weekly performance reviews ├── Monthly optimization updates ├── Quarterly capability expansion ├── Continuous guardrail refinement └── Regular training and updates Deliverables: ├── Daily/weekly/monthly reports ├── Optimization recommendations ├── Capability roadmap ├── Training materials └── Continuous improvement plan ``` --- ## Performance Metrics & Monitoring ### Key Performance Indicators | Metric | Target | Purpose | |---|---|---| | Resolution Rate | 50-70% (first contact) | Measure autonomous handling | | Escalation Rate | <30% | Control human workload | | Customer Satisfaction (CSAT) | >80% | Measure experience quality | | Average Resolution Time | -30% vs human | Track efficiency | | Task Completion Rate | >85% | Measure task success | | Knowledge Accuracy | >95% | Ensure correct information | | Policy Adherence | 100% | Compliance verification | | Guardrail Violations | <0.1% | Safety monitoring | ### Real-Time Monitoring Dashboard ``` Agentic Virtual Agent Monitor (Live) Queue: Customer Service ├── Active Interactions: 47 ├── Agents (agentic): 3 active ├── Human Agents: 12 available │ ├── Current Metrics: │ ├─ Avg Resolution: 4.2 minutes │ ├─ Current CSAT: 4.3/5 (sample) │ ├─ Escalation Rate: 22% │ └─ Policy Adherence: 100% │ ├── Interaction Breakdown: │ ├─ Self-service (no agent): 1,250 today │ ├─ Agentic agent resolved: 480 today │ ├─ Escalated to human: 85 today │ └─ In-progress: 47 │ ├── Top Use Cases: │ ├─ Order Status (156 resolved) │ ├─ Billing Questions (134 resolved) │ └─ Returns (89 resolved) │ └── Guardrail Status: ├─ Critical Rules: All passing ├─ Violations Today: 0 └─ Approval Requests: 3 pending ``` --- ## Common Implementation Scenarios ### Scenario 1: Customer Service Center Automation ``` Organization: E-commerce Company (100 agents) Current: High volume, repetitive inquiries Deployment: ├─ Agentic agents for: Order status, returns, │ exchanges, tracking, simple complaints ├─ Resolution targets: 60-70% automation ├─ Channels: Chat, email, phone └─ Timeline: 10 weeks Expected Results: ├─ Automation Rate: 65% of volume ├─ Agent Productivity: +40% (less routine work) ├─ CSAT: +12% (faster resolution) ├─ AHT: -25% (focused on complex issues) ├─ Cost Reduction: $180K-250K annually └─ Payback Period: 4-5 months ``` ### Scenario 2: Technical Support Evolution ``` Organization: SaaS Company (50 support agents) Current: Mixed simple and complex tickets Deployment: ├─ Agentic agents for: FAQ, basic troubleshooting, │ account resets, documentation lookup ├─ Resolution targets: 40-50% automation ├─ Channels: Chat, email, knowledge portal └─ Timeline: 12 weeks Expected Results: ├─ Automation Rate: 45% of volume ├─ Agent Capacity: +30% (handling complex) ├─ CSAT: +8% (better first contact) ├─ Time to Resolution: -20% ├─ Escalation Quality: Improved (richer context) └─ Cost Reduction: $120K-150K annually ``` ### Scenario 3: Billing & Collections Department ``` Organization: Financial Services (30 agents) Current: Payment processing, dispute resolution Deployment: ├─ Agentic agents for: Payment processing, │ arrangement setup, billing inquiries, │ promotional application ├─ Resolution targets: 55-65% automation ├─ Channels: Voice, chat, IVR └─ Timeline: 14 weeks Expected Results: ├─ Automation Rate: 60% of volume ├─ First-Contact Collections: +25% ├─ Customer Payment Satisfaction: +15% ├─ Dispute Resolution: 50% autonomous ├─ Revenue Impact: +$50K monthly └─ Payback Period: 3-4 months ``` --- ## Best Practices for Agentic Virtual Agents ### Agent Design - **Clear Boundaries** - Define exactly what agent can and cannot do - **Progressive Complexity** - Start simple, add complexity gradually - **Natural Conversation** - Design for human-like interaction - **Graceful Degradation** - Handle misunderstandings smoothly - **Escalation Clarity** - Know when to hand off to human - **Continuous Learning** - Improve based on interaction outcomes - **Regular Updates** - Refine agent behavior based on data ### Safety & Governance - **Guardrail Testing** - Rigorously test all safety mechanisms - **Approval Workflows** - Require approval for high-risk actions - **Audit Trails** - Log all agent decisions and actions - **Compliance Monitoring** - Ensure all interactions meet requirements - **Regular Audits** - Review agent behavior periodically - **Exception Handling** - Clear process for policy exceptions - **Transparency** - Customers know they're talking to agent ### Integration Management - **System Reliability** - Ensure backend systems are available - **Data Quality** - Keep knowledge base and data current - **Error Handling** - Gracefully handle system failures - **Transaction Safety** - Verify critical operations succeed - **Real-time Connectivity** - Fast system access during interactions - **Fallback Plans** - Handle integration failures smoothly - **Performance Optimization** - Keep response times fast ### Performance & Optimization - **Monitor Continuously** - Track metrics in real-time - **Analyze Failures** - Understand why escalations occur - **Gather Feedback** - Customer and agent feedback critical - **Iterate Quickly** - Make improvements based on data - **A/B Testing** - Test different approaches - **Seasonal Planning** - Adjust for peak periods - **Capacity Planning** - Scale agents with demand --- ## Troubleshooting Common Issues | Issue | Cause | Resolution | |---|---|---| | High escalation rate | Agent lacks capability for use case | Expand agent training and knowledge | | Poor customer satisfaction | Conversation feels unnatural | Refine language generation and tone | | Integration failures | Backend system issues | Test integrations, improve error handling | | Slow response times | System latency or complex reasoning | Optimize queries, simplify logic | | Guardrail violations | Rules too loose or unclear | Tighten rules, improve monitoring | | Incorrect decisions | Knowledge gaps or logic errors | Update knowledge base, refine rules | | Customers can't escalate | Escalation path unclear | Add clear escalation triggers | | Token overages | Agent using more tokens than expected | Optimize agent logic, reduce complexity | | Integration data stale | Knowledge base not updated | Establish content review process | | Customer confusion | Unclear communications | Simplify language, improve clarity | --- ## Agentic Capabilities Evolution ### Current State (2026) ``` What Agentic Agents Can Do Now: ├─ Understand complex customer requests ├─ Reason through multi-step problems ├─ Access and apply knowledge ├─ Execute transactions within bounds ├─ Maintain context across conversation ├─ Adapt behavior based on customer ├─ Learn from outcomes └─ Handle 50-70% of customer interactions ``` ### Near-term (2026-2027) ``` Expected Enhancements: ├─ Deeper business system integration ├─ More sophisticated reasoning ├─ Better multi-language support ├─ Improved emotion detection ├─ Proactive outreach capabilities ├─ Cross-department orchestration └─ Enhanced learning algorithms ``` ### Future Vision (2027+) ``` Potential Capabilities: ├─ Full autonomy within broad boundaries ├─ Predictive problem prevention ├─ Seamless cross-organization collaboration ├─ Personalized experience generation ├─ Advanced negotiation capability ├─ Complex financial decisions └─ Level 5 - Fully autonomous agents ``` --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What makes agentic agents different? | They reason and plan solutions rather than following fixed scripts | | What level of autonomy do they have? | Level 4 - semi-autonomous with defined guardrails and boundaries | | What issues can they handle? | 50-70% of customer interactions, including complex scenarios | | How do guardrails work? | Configurable rules that define what agents can/cannot do | | Can they access backend systems? | Yes, can execute transactions and retrieve real-time data | | How do they learn? | Analyze interaction outcomes and improve responses over time | | What happens if they can't solve? | Intelligent escalation to human with full context | | How long to deploy? | 8-14 weeks depending on complexity | | What's the expected ROI? | 4-6 months through automation and efficiency gains | | What channels do they support? | All Genesys channels - voice, chat, email, messaging, social | | How much does it cost? | AI Experience tokens metered by usage | | What about compliance? | Built-in audit trails, guardrails, and monitoring | | Can they make mistakes? | Yes - monitored and corrected through guardrails | | Are humans always available? | Yes - escalation available anytime | | What's most important for success? | Clear use cases, good data, proper guardrails | --- ## Key Takeaways - **Intelligent Autonomy** - Agentic agents reason and plan solutions independently - **Safety by Design** - Configurable guardrails ensure compliance and safety - **Significant Automation** - Handle 50-70% of customer interactions without human - **Omnichannel Ready** - Work seamlessly across all communication channels - **Continuous Learning** - Improve performance based on interaction outcomes - **Graceful Escalation** - Intelligently hand off to humans when needed - **Fast Deployment** - Weeks vs. months to implement complex automation - **Strong ROI** - Typical payback in 4-6 months - **Customer Focused** - Natural conversation, context-aware, personalized - **Future Ready** - Path to higher autonomy levels as technology matures --- ## Real-World Success Metrics ### Customer Service Example ``` Before Agentic Agent: ├─ Customer Resolution Rate: 65% (human agents) ├─ Average Handle Time: 8 minutes ├─ Cost per Interaction: $4.50 ├─ Customer Satisfaction: 75% └─ Monthly Volume: 10,000 interactions After Agentic Agent Implementation: ├─ Self-service + Agent: 80% Resolution Rate ├─ Agentic Agent Only: 65% first contact ├─ Average Handle Time: 3.5 minutes ├─ Cost per Interaction: $1.80 ├─ Customer Satisfaction: 82% └─ Same Volume but 40% labor reduction Annual Impact: ├─ Cost Savings: ~$180,000/year ├─ Improvement in CSAT: +7 points ├─ Faster Resolution: -55% time └─ ROI: 350% in first year ``` ### Financial Services Example ``` Before: ├─ Payment Processing: Human handled 70% ├─ Average Call Time: 12 minutes ├─ Collections Rate: 82% ├─ Cost per Transaction: $6.00 └─ Agent Utilization: 85% After Agentic Implementation: ├─ Agentic Agent: 60% self-service ├─ Human Agents: 40% complex cases ├─ Average Call Time: 4 minutes ├─ Collections Rate: 87% ├─ Cost per Transaction: $2.40 ├─ Agent Utilization: 65% (more valuable work) Impact: ├─ Cost Savings: $240K/year ├─ Collections Improvement: +5% ├─ Revenue Benefit: $150K/year └─ Total Benefit: $390K/year ``` --- ## Getting Started Checklist ### Assessment Phase - [ ] Audit current agent performance - [ ] Identify top 3-5 use cases - [ ] Assess customer journey complexity - [ ] Define success metrics - [ ] Analyze integration requirements - [ ] Document guardrail needs - [ ] Assess team readiness ### Planning Phase - [ ] Prioritize use cases by complexity - [ ] Create implementation roadmap - [ ] Design guardrails and boundaries - [ ] Plan change management - [ ] Allocate resources - [ ] Set realistic timelines - [ ] Establish success criteria ### Development Phase - [ ] Build agentic agents - [ ] Configure knowledge base - [ ] Set up integrations - [ ] Configure guardrails - [ ] Create monitoring dashboards - [ ] Document processes - [ ] Prepare testing plan ### Deployment Phase - [ ] Conduct thorough testing - [ ] Deploy to pilot - [ ] Monitor closely - [ ] Gather feedback - [ ] Optimize configuration - [ ] Full production rollout - [ ] Establish support procedures ### Optimization Phase - [ ] Monitor daily metrics - [ ] Analyze performance - [ ] Gather customer feedback - [ ] Implement improvements - [ ] Scale gradually - [ ] Plan next use cases - [ ] Document learnings --- ## Additional Resources ### Official Documentation Links - Virtual Agent Overview: help.genesys.cloud/articles/virtual-agent-overview/ - AI Studio & AI Guides: help.genesys.cloud/articles/about-ai-studio/ - Agentic Capabilities: help.genesys.cloud/articles/agentic-virtual-agents/ - Architect Virtual Agent Flows: help.genesys.cloud/articles/architect-virtual-agent-flows/ ### Training & Support - Genesys University: genesys.com/training - Community Forums: https://community.genesys.com - Technical Support: https://support.genesys.com - Sales Support: sales@genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation **Based On:** AI Studio, Virtual Agent, and Agentic Capabilities releases **Version:** 1.0 # Predictive Routing # Genesys PureCloud Predictive Routing Documentation ## Study Notes | Topic | Description | |---|---| | Predictive Routing | Machine learning-powered intelligent routing that matches contacts to optimal agents | | AI Engine | White-box models with explainability features showing feature importance scores | | Outcome Labels | Custom KPI metrics tracked and used to retrain models continuously | | Queue Configuration | Per-queue settings enabling/disabling predictive routing with KPI selection | | Model Retraining | Weekly automated retraining with daily data updates ensuring current accuracy | | Data Retention | No model data retained beyond 90 days for privacy and compliance | --- ## Navigation Admin → Architect → Routing → Predictive Routing Configuration OR Admin → Contact Center → Routing → Predictive Routing Settings --- ## Predictive Routing Overview Genesys Cloud's predictive routing uses machine learning to rank agents for optimal handling of interactions, supporting key performance indicators like average handle time and next contact avoidance. Unlike traditional rule-based routing, predictive routing analyzes hundreds of data points in real-time to identify which agent is most likely to deliver the best customer outcome based on your defined business goals. ### Key Capabilities - Predictive routing uses white-box models that allow gaining insights into how the features contribute to a prediction, with each input feature given a percentage/score that represents its importance - Continuously improves scoring accuracy based on outcome data from previous interaction-agent matchups - Real-time agent scoring based on historical performance and current state - Automatically retrains and updates features used for agent scoring with daily data and retrains the data models weekly - Support for custom KPI optimization (AHT, NCA, CSAT, etc.) - Automatic fallback when models are unavailable ### How It Works 1. Contact arrives at system 2. Contact metadata extracted (type, queue, customer data) 3. URS Strategy Subroutines submit interaction details to the Core Platform, which scores agents for their historical ability to handle such an interaction 4. Machine learning model scores all available agents 5. Contact routed to highest-scoring agent 6. Interaction outcome captured and logged 7. Models updated with new outcome data for future predictions --- ## Edition & Module Requirements | Requirement | Details | |---|---| | Minimum Edition | Premium Edition (Genesys Cloud CX 1-4) | | Module | Workforce Optimization or Predictive Routing add-on | | License Type | Included with qualifying editions | | AI Tokens | May consume tokens depending on configuration | | Setup | Admin access to Architect and routing configuration | --- ## Study Notes - AI Model Features | Feature Category | Examples | Impact | |---|---|---| | Agent Profile Data | Skills, tenure, department, certifications | High - Core matching criteria | | Agent Performance Data | Historic AHT, FCR, quality scores, tenure | High - Predicts future performance | | Availability Data | Current status, workload, after-call-work | High - Real-time readiness | | Interaction Metadata | Contact type, channel, customer segment | Medium - Context matching | | Queue Statistics | Historical patterns, time-of-day trends | Medium - Volume prediction | | Customer Data | History, value, previous interactions | Low-Medium - Personalization | --- ## AI Model Architecture ### White-Box Model Explainability Genesys helps you deduce the prediction by presenting a global interpretation that describes the average behavior of a model. A high value means that the feature will have a larger effect on the model's predictions and ranking of agents. While a small value is given to unimportant features whose contribution is mostly ignored for the model's predictions. ### Model Training & Retraining To keep up with changing levels of agent proficiency and customer interaction contexts, the data models continually retrain and learn from the latest features. Genesys Cloud updates the features used for agent scoring with daily data and retrains the data models weekly. No data is retained in the models for more than 90 days. ### Privacy & Compliance Genesys does not use PII for the agent scoring process. Genesys Cloud only uses transaction conversation data to train machine learning models. --- ## Data Requirements for Optimal Performance ### Recommended Data Volume Genesys recommends a periodic upload of outcome data to ensure better model training and prediction accuracy. The recommended frequency is once a day, or, at the minimum, once a week. Genesys recommends at least 90 days of data and ideally 180 days of data. If you retain fewer days of data, you can still use and benefit from predictive routing. However, the quality and effectiveness of your AI models and predictions is likely to suffer and the resulting benefit will be reduced compared to standard routing. ### Data Sources - Agent profile data such as skills, tenure, department, certificates, employee type - Agent performance data such as historic average handle time for a queue - Custom outcome data via Outcome Attributions API - External data sources for outcome-based KPIs ### Minimum Data Thresholds Data, even if available, is considered for the purpose of routing calculation only if the data volume meets a minimum requirement. If the volume of data does not meet the specified threshold value, the machine learning model does not use the available data. While not mandatory for Genesys predictive routing to operate, for better prediction results, Genesys recommends that you ensure the availability of sufficient volume of data against maximum number of fields. --- ## Outcome Labels & Custom KPIs ### Default KPIs Predictive Routing can optimize for standard Genesys metrics: - **Average Handle Time (AHT)** - Minimize time to resolve - **Next Contact Avoidance (NCA)** - Minimize customer repeat contacts - **Customer Satisfaction (CSAT)** - Maximize customer satisfaction scores - **First Contact Resolution (FCR)** - Prevent escalations and transfers ### Custom Outcome Labels If you use outcome-based custom KPIs, Genesys predictive routing relies on data from external data sources, which it receives through the Outcome Attributions API. ### Uploading Outcome Data - **Frequency**: Daily recommended, weekly minimum - **Format**: CSV or API integration via Outcome Attributions API - **Data Requirements**: At least 90 days, ideally 180 days for optimal models - **Processing**: Data uploaded to Genesys Cloud for model retraining - **Model Updates**: Weekly retraining cycle incorporates latest outcomes ### Setting KPI Optimization per Queue During queue configuration, administrators select which KPI the predictive routing model should optimize for that specific queue. The system will then score agents based on their historical performance against that metric. --- ## Queue Configuration Steps ### Step 1: Navigate to Routing Settings 1. Log into Genesys Cloud Admin 2. Go to Architect → Routing 3. Select queue to enable predictive routing 4. Open queue configuration settings ### Step 2: Enable Predictive Routing 1. Locate "Predictive Routing" toggle 2. Toggle ON to enable feature 3. System validates that requirements are met 4. Configuration page appears ### Step 3: Select KPI Optimization 1. Choose primary KPI from dropdown: - Average Handle Time (AHT) - Next Contact Avoidance (NCA) - Customer Satisfaction (CSAT) - Custom outcome metric 2. If custom KPI: Verify outcome data is being uploaded via API 3. Save selection ### Step 4: Configure Scoring Parameters 1. Set minimum agent availability threshold 2. Define required skills for queue 3. Configure fallback routing behavior 4. Set scoring timeout (default: 3 seconds) 5. Enable/disable explainability logging ### Step 5: Data & Outcome Mapping 1. Confirm queue is sending outcome data 2. For outcome-based custom KPIs, ensure periodic upload of outcome data via Outcome Attributions API 3. Map external outcome labels to routing KPI 4. Verify historical data volume meets minimums 5. Test data flow ### Step 6: Validation & Testing 1. Enable predictive routing on small percentage of traffic 2. Monitor model accuracy metrics 3. Verify fallback routing works correctly 4. Gather baseline performance data 5. Review feature importance scores in reports ### Step 7: Gradual Rollout 1. Increase traffic percentage gradually 2. Monitor agent utilization and contact distribution 3. Track KPI impact daily 4. Make refinements as needed 5. Scale to 100% when confident in performance --- ## Model Scoring Process ### Agent Scoring Flow ``` Incoming Contact ↓ Extract Contact Metadata ├── Contact type (voice, chat, email) ├── Queue assignment ├── Customer segment └── Interaction intent ↓ Query AI Model ├── Input: Contact metadata ├── Features: Agent data, performance, availability ├── Process: ML scoring algorithm └── Output: Agent scores (0-100) ↓ Rank Agents ├── Agent A: 87 score ├── Agent B: 72 score ├── Agent C: 91 score (highest) └── Agent D: 65 score ↓ Route to Highest Scorer ├── Select: Agent C (91 score) ├── Fallback: If unavailable → Agent A (87) └── Escalation: If all unavailable → Queue ↓ Log Interaction ├── Selected agent score ├── Feature importance values ├── Outcome when complete └── Update training data ``` ### Feature Importance Explanation The model provides transparency showing which factors most influenced each routing decision: - **High Importance Features** (weight >10%) - Directly impact agent selection - Should be monitored and optimized - **Medium Importance Features** (weight 5-10%) - Contribute meaningfully to decisions - Worth tracking for insights - **Low Importance Features** (weight <5%) - Minimal impact on routing - May indicate data quality issues --- ## Real-World Implementation Scenario ### Banking Contact Center ``` Queue: Mortgage Support KPI Optimization: Average Handle Time (AHT) Agent Profiles: ├─ Agent Sarah │ ├─ Skills: Mortgage, Refinance, Loan Modification │ ├─ Avg AHT: 6.2 minutes │ ├─ Tenure: 5 years │ └─ Current: Available (0 contacts) │ ├─ Agent James │ ├─ Skills: Mortgage, Basic Support │ ├─ Avg AHT: 8.1 minutes │ ├─ Tenure: 1 year │ └─ Current: Available (1 contact) │ └─ Agent Maria ├─ Skills: Mortgage, Refinance, Advanced ├─ Avg AHT: 5.9 minutes ├─ Tenure: 7 years └─ Current: Available (2 contacts) Incoming Contact: "I'd like information about refinancing my mortgage" Model Analysis: ├─ Contact Intent: Refinance inquiry ├─ Expected AHT: ~7 minutes ├─ Required Skills: Refinance, Mortgage └─ Channel: Voice Agent Scoring (for AHT optimization): ├─ Sarah: 89 score │ └─ Reasoning: Strong refinance skills, excellent AHT, │ lower current load ├─ James: 71 score │ └─ Reasoning: Has required skills but higher baseline AHT └─ Maria: 85 score └─ Reasoning: Best AHT historically but already has higher load (2 contacts) Routing Decision: Route to Agent Sarah (highest score: 89) Expected Outcome: ├─ Faster resolution (Sarah's AHT advantage) ├─ Better customer experience ├─ Optimizes against AHT KPI └─ Load distributed effectively ``` --- ## Model Metrics & Dashboard ### Key Performance Indicators Tracked | Metric | Purpose | Good Range | |---|---|---| | Model Accuracy | How often predictions are correct | >75% | | Feature Coverage | % of required data available | >85% | | Agent Utilization | Even work distribution | 70-90% | | KPI Improvement | Impact on selected metric | +5-20% | | Fallback Rate | When model can't score | <5% | ### Monitoring Model Health - **Daily**: Check model score distribution - **Weekly**: Review prediction accuracy vs. actual outcomes - **Monthly**: Analyze feature importance changes - **Quarterly**: Assess KPI impact and ROI --- ## Best Practices ### Data Quality - **Validate Outcome Data** - Ensure outcome labels are accurate and complete - **Consistent Data Format** - Maintain standardized data formats in API uploads - **Timely Uploads** - Daily uploads ensure models have latest information - **Complete Records** - Capture outcomes for all interactions, not just subset ### Model Optimization - **Start with One KPI** - Master one optimization goal before multiple - **Monitor Feature Changes** - Track how feature importance shifts over time - **A/B Test Approaches** - Compare predictive routing vs. traditional on test queues - **Iterative Improvement** - Refine based on real-world performance data ### Queue Configuration - **Gradual Rollout** - Start with 10% traffic, increase to 100% gradually - **Skill Validation** - Ensure agent skill assignments are accurate and current - **Outcome Mapping** - Correctly map external KPIs to routing selection - **Fallback Testing** - Verify routing works when models unavailable ### Change Management - **Communicate Changes** - Inform agents about predictive routing changes - **Monitor Agent Sentiment** - Track agent acceptance and satisfaction - **Provide Training** - Explain how predictive routing affects their work - **Set Expectations** - Clear goals and targets for improvement --- ## Common Implementation Issues & Solutions | Issue | Cause | Solution | |---|---|---| | Model not scoring agents | Insufficient data volume | Ensure 90+ days of outcome data available | | Uneven agent utilization | Agents with similar skills | Review and update skill assignments | | No KPI improvement | Wrong KPI selected for queue | Validate KPI choice aligns with business goals | | High fallback rate | Skill mismatches in data | Audit and correct agent skill inventory | | Slow model updates | Outcome data not uploading | Verify API integration and data flow | | Feature importance unclear | New queue without history | Wait for sufficient data accumulation | | Model accuracy low | Poor quality training data | Validate and clean outcome data | --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What is Predictive Routing? | ML-powered routing that matches contacts to optimal agents based on historical data | | What models does it use? | White-box models with explainability showing feature importance percentages | | How often do models retrain? | Weekly with daily data updates; no data retained beyond 90 days | | What data is needed? | 90+ days recommended (ideally 180); daily or weekly outcome uploads | | What KPIs can it optimize? | AHT, NCA, CSAT, or custom metrics via Outcome Attributions API | | Where do you configure it? | Admin → Architect → Routing → Queue settings | | Does it use PII? | No, only transaction conversation data for training | | What's the expected improvement? | 5-20% improvement in selected KPI | | How do you upload outcomes? | Via Outcome Attributions API or CSV upload to Genesys Cloud | | What if model unavailable? | Falls back to traditional routing automatically | | How do agents get selected? | Highest-scoring agent routed first; fallback to next highest if unavailable | | Can you explain routing decisions? | Yes, white-box models provide feature importance scores for transparency | --- ## Key Takeaways - **Intelligent Matching** - By continuously learning from real-time and historical data, it helps optimize important KPIs like average handle time, first-contact resolution (through next contact avoidance (NCA)) and more - **White-Box Transparency** - Models explain exactly which factors influenced each routing decision - **Continuous Learning** - Weekly retraining with daily updates ensures models stay current - **Privacy-First** - No PII used; data retained only 90 days - **Custom Optimization** - Select any KPI (standard or custom) to optimize routing around - **Data-Driven** - Requires quality outcome data for accurate predictions - **Graceful Fallback** - Traditional routing if model unavailable - **Per-Queue Configuration** - Each queue can optimize for different KPIs - **Explainability** - Feature importance scores show transparency in AI decisions - **Proven ROI** - 5-20% improvement typical on selected KPI --- ## Additional Resources ### Official Documentation Links - Predictive Routing Overview: help.genesys.cloud/articles/predictive-routing-overview/ - AI Model Scoring: help.genesys.cloud/articles/how-the-ai-model-scores-agents-for-predictive-routing/ - Data Requirements: help.genesys.cloud/articles/sources-of-data-for-predictive-routing-decisions/ - Use of AI in Routing: help.genesys.cloud/articles/use-of-ai-in-predictive-routing/ ### Support & Training - Genesys University: genesys.com/training - Community Forums: https://community.genesys.com - Technical Support: https://support.genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation (help.genesys.cloud) **Validated:** Current with January-March 2026 releases **Version:** 1.0 # Agent Copilot # Genesys PureCloud Agent Copilot Documentation ## Study Notes | Topic | Description | |---|---| | Agent Copilot | Real-time AI assistance for agents during customer interactions | | Core Function | Determines customer intent and provides next best actions in real-time | | After-Call Work | Automates summaries, wrap-up codes, and resolution tracking | | AI Skills | Modular AI capabilities that can be combined and customized | | On-Queue Behavior | Operates continuously during interaction, adapting to conversation flow | | Key Results | 5-min AHT reduction, 1.5-min hold time reduction, 2-min ACW reduction | --- ## Navigation Admin → Contact Center → Agent Assistance → Agent Copilot Configuration OR Admin → AI Studio → Agent Copilot Settings --- ## Agent Copilot Overview Genesys Agent Copilot enhances communication between the agent and the customer by determining customer intent and providing relevant next best actions to the agent. It offers assistance in after-call work and provides a summary of the conversation, reason for contact, resolution, and suggests wrap-up codes. Genesys Agent Copilot integrates AI and Large Language Models to deliver intelligent assistance capabilities that transform agent productivity and customer service delivery. The platform automates after-call work through intelligent wrap-up code suggestions, conversation summaries, and resolution tracking, while providing real-time next-best-action recommendations and customer intent determination during live interactions. ### Key Capabilities - Real-time agent assist during customer interactions on voice and digital channels - Intelligent next-best-action recommendations based on conversation context - Automatic conversation summaries and wrap-up code suggestions - Customer intent determination and understanding - Multi-language support (13+ languages including English, Spanish, French, German, Japanese, Korean, Arabic, Hindi, Portuguese, Swedish, Dutch, Italian) - Custom AI Guides built through AI Studio integration - Knowledge article integration and answer highlighting - Queue-specific configuration and management - Performance analytics and monitoring dashboards ### Performance Metrics Agent Copilot users have seen results including: - 5-minute decrease of average handle time - 1.5-minute decrease in average hold time - 2-minute decrease in after-call work time --- ## Edition & Module Requirements | Requirement | Details | |---|---| | Minimum Edition | Premium Edition (Genesys Cloud CX 1-4) | | Module | Included or Agent Copilot add-on depending on tier | | License Type | Per-user licensing | | AI Tokens | Requires tokens: 40-60 per user for Agent Copilot | | Knowledge Base | Genesys Knowledge Workbench or Knowledge Fabric recommended | --- ## Study Notes - Agent Copilot Features | Feature | Description | Benefit | |---|---|---| | Intent Recognition | AI understands customer goals from conversation | Faster issue identification | | Next-Best-Action | Recommends optimal next step for agent | Guides resolution path | | Auto-Summarization | Generates conversation summary automatically | Reduces after-call work | | Wrap-Up Code Suggestions | AI suggests appropriate completion codes | Faster code selection | | Knowledge Integration | Surfaces relevant knowledge articles automatically | Faster access to information | | Answer Highlighting | Highlights key parts of knowledge articles | Easier information scanning | | Sentiment Monitoring | Detects customer emotion in real-time | Enables de-escalation | | Multi-Language Support | Operates in 13+ languages | Global team support | | Custom Guides | Build AI agents using AI Studio | Automation flexibility | | Performance Analytics | Dashboard showing copilot usage and impact | Measurement and optimization | --- ## How Agent Copilot Works ### During Interaction (Real-Time Assistance) ``` Agent Answers Contact ↓ Copilot Begins Monitoring ├── Real-time transcription ├── Intent analysis └── Context understanding ↓ Customer Explains Issue "I'm having trouble logging into my account" ↓ Copilot Analyzes ├── Intent: Account Access Issue ├── Context: Login problem ├── Emotion: Slightly frustrated └── Knowledge needed: Account recovery ↓ Copilot Provides Assistance ├── Knowledge Articles │ ├─ "Password Reset Steps" (94% match) │ ├─ "Two-Factor Auth Help" (81% match) │ └─ "Account Locked Recovery" (76% match) │ ├── Next-Best-Action │ └─ "Guide customer through password reset" │ └── Script Suggestion └─ "I can help you regain access to your account..." ↓ Agent Reviews & Applies ├── Reads suggested article ├── Guides customer through steps ├── Issue resolved └── Interaction continues ``` ### After Interaction (After-Call Work) ``` Contact Completes ↓ Copilot Generates Summary ├── Conversation Analysis ├── Key Points Extraction ├── Issue Resolution Status └── Customer Sentiment Analysis ↓ Summary Displayed to Agent "Customer called about password reset. Issue resolved through email link sent. Customer satisfied." ↓ Wrap-Up Code Suggestions ├─ Suggested Codes (in priority order) │ ├─ Password Reset (89% confidence) │ ├─ Account Access (71% confidence) │ └─ Technical Support (34% confidence) └─ Agent selects from suggestions ↓ Agent Reviews & Saves ├── Reviews summary (can edit) ├── Selects wrap-up code ├── Adds notes if needed └── Saves interaction ``` --- ## AI Skills Architecture AI Skills are modular, customizable AI capabilities that can be combined and deployed through Agent Copilot. Each skill targets a specific function or use case. ### Types of AI Skills **Knowledge Skills** - Surfaces relevant knowledge articles automatically - Answers customer questions from knowledge base - Answer highlighting for quick scanning - Knowledge caching for speed **Next-Best-Action Skills** - Recommends optimal next step in interaction - Guides agent through proper procedure - Suggests escalation or transfer when needed - Predicts customer outcomes **Wrap-Up Code Skills** - Analyzes interaction to suggest wrap-up codes - Reduces time agents spend on code selection - Improves code accuracy through AI - Learning improves accuracy over time **Sentiment & De-Escalation Skills** - Real-time sentiment detection - De-escalation recommendations - Tone guidance for responses - Emotional intelligence coaching **Compliance & Risk Skills** - Monitors for compliance violations - Alerts on risky language or actions - Ensures required disclosures made - Flags suspicious patterns **Custom Business Skills** - Created via AI Studio - Built using AI Guides - Business-specific logic - Adaptive to business processes --- ## On-Queue Behavior ### How Agent Copilot Operates During Customer Interaction ``` Queue: Customer Support Status: Agent Sarah actively handling call Real-Time Copilot Behavior: MOMENT 1 (0:15 into call) ├─ Copilot listening to conversation ├─ Transcribing in real-time ├─ Analyzing customer statements └─ Building context understanding MOMENT 2 (0:45 into call) Customer: "I need to change my billing address" ├─ Intent Detected: Billing Update ├─ Copilot triggers knowledge search ├─ Results: 2 relevant articles found └─ Displayed to Agent Sarah MOMENT 3 (1:15 into call) ├─ Customer emotion: Neutral → Positive ├─ Issue complexity: Routine ├─ Next action: Process address change └─ Copilot suggests: "Enter new address in system" MOMENT 4 (1:45 into call) Customer: "Can I also update my phone number?" ├─ New intent added to context ├─ Copilot expands knowledge articles ├─ Related articles surfaced └─ Agent handles both requests together MOMENT 5 (3:00 - Call ends) ├─ Both issues resolved ├─ Customer satisfied ├─ Copilot prepares summary └─ Ready for after-call work ``` ### Agent Copilot Interface Components ``` Agent Desktop View ┌──────────────────────────────────────┐ │ ACTIVE CALL - Sarah Martinez │ │ Customer: John Smith │ │ Queue: Billing Support │ │ Duration: 3:42 │ └──────────────────────────────────────┘ ┌──────────────────────────────────────┐ │ 📚 SUGGESTED KNOWLEDGE │ ├──────────────────────────────────────┤ │ ✓ Update Billing Address (94%) │ │ ✓ Change Phone Number (87%) │ │ ○ Payment Methods (61%) │ │ │ │ [Show Full Articles] [Minimize] │ └──────────────────────────────────────┘ ┌──────────────────────────────────────┐ │ ➡️ NEXT-BEST-ACTION │ ├──────────────────────────────────────┤ │ Recommended: Process address update │ │ in customer record │ │ │ │ Steps: │ │ 1. Confirm new address │ │ 2. Verify zip code │ │ 3. Update system │ │ 4. Send confirmation │ └──────────────────────────────────────┘ ┌──────────────────────────────────────┐ │ 😊 CUSTOMER SENTIMENT │ ├──────────────────────────────────────┤ │ Current: POSITIVE │ │ Emotion: Satisfied │ │ Recommendation: Offer additional help│ └──────────────────────────────────────┘ [Notes Box] [Hold/Transfer/Close Buttons] ``` ### Behavior by Interaction Type **Inbound Call** - Continuous listening and transcription - Real-time intent and sentiment analysis - Knowledge surfacing as topics emerge - Next-best-action recommendations throughout - Escalation alerts if conversation flags - Summary generation at call end **Chat/Email** - Message-by-message analysis - Delayed but thorough knowledge search - Context building across multiple messages - Suggested responses for agent review - Quick-reply templates with personalization - Handoff summaries for escalation **Outbound Call** - Prepopulates customer context - Suggests opening statements - Guides conversation flow - Handles objections with recommendations - Closing suggestions - Outcome tracking --- ## Implementation Guide ### Step 1: Prerequisites & Planning 1. Ensure Premium Edition activated 2. Confirm sufficient AI tokens available (40-60 per user) 3. Audit knowledge base quality and coverage 4. Identify queues for initial deployment 5. Assess agent readiness and training needs 6. Plan change management approach ### Step 2: Knowledge Base Setup 1. Review/create knowledge articles in Knowledge Workbench 2. Ensure articles are accurate and current 3. Add metadata and keywords for searchability 4. Organize by topic and queue 5. Set article access permissions 6. Test knowledge base connectivity ### Step 3: Enable Agent Copilot 1. Navigate to Admin → Contact Center → Agent Assistance 2. Select "Agent Copilot" → Enable 3. Choose knowledge source (Knowledge Workbench v2 or Fabric) 4. Configure recommendation parameters 5. Set recommendation types to display 6. Choose display behavior (auto-popup vs. agent-triggered) ### Step 4: Configure by Queue 1. Create queue-specific Copilot settings 2. Define which AI Skills are active per queue 3. Configure knowledge sources 4. Set wrap-up code matching rules 5. Establish sentiment thresholds 6. Test queue-specific behavior ### Step 5: Agent Training 1. Conduct overview training on Copilot interface 2. Demonstrate real-time knowledge suggestions 3. Practice reviewing and applying recommendations 4. Explain wrap-up code suggestions 5. Cover sentiment alerts and de-escalation 6. Q&A and feedback ### Step 6: Phased Rollout 1. Deploy to pilot queue first 2. Monitor closely for 1-2 weeks 3. Gather agent and supervisor feedback 4. Optimize based on learnings 5. Expand to additional queues 6. Scale to full deployment ### Step 7: Ongoing Monitoring 1. Review Agent Copilot dashboard daily 2. Track key metrics (AHT reduction, token usage) 3. Monitor agent adoption rates 4. Gather continuous feedback 5. Refine knowledge articles based on usage 6. Optimize AI recommendations --- ## Real-Flow Scenarios ### Scenario 1: Technical Support with Knowledge Integration ``` Agent: "Thanks for calling technical support, how can I help?" Customer: "My printer isn't connecting to WiFi" Copilot immediately surfaces: ├─ 3 relevant knowledge articles ├─ Troubleshooting flowchart visual ├─ Video walkthrough link └─ Quick-resolution steps Agent: "I can help. Let me walk you through the WiFi connection steps which usually resolve this..." [Uses Copilot-suggested steps to guide customer] Result: Issue resolved in 4 minutes Copilot Summary: "Customer called about printer WiFi connectivity issue. Guided through troubleshooting steps. Issue resolved after resetting router. Customer satisfied." Wrap-up Code Suggestions: ├─ Printer Setup (92% confidence) ← Selected ├─ Technical Support (67% confidence) └─ Network Issues (45% confidence) Time Saved: 2 minutes (no manual summary or code lookup) ``` ### Scenario 2: Billing with Sentiment De-escalation ``` Customer (frustrated): "Why was I charged twice?!" Copilot detects: ├─ Emotion: FRUSTRATED ├─ Sentiment score: -2.1/5 ├─ Recommended action: Empathize & resolve immediately └─ De-escalation tip: "Acknowledge frustration sincerely" Agent (applying suggestion): "I completely understand your frustration. Let me look into that right away and fix this for you." Copilot provides: ├─ Knowledge: "Duplicate Charge Resolution" ├─ Next action: "Issue refund immediately" └─ Script: "Here's what happened and how I'm fixing it..." [Agent processes refund while explaining] Customer (relieved): "Oh wow, thanks for handling that so fast!" Copilot updates sentiment: +1.5/5 (positive) Summary generated: "Customer called upset about duplicate charge. Explained system error, issued refund immediately. Customer very satisfied with quick resolution and empathetic service." Result: De-escalation successful, issue resolved, positive outcome ``` ### Scenario 3: Sales with Cross-Sell Recommendations ``` Customer: "I'd like to upgrade my plan" Copilot analyzes: ├─ Customer history: 2-year subscriber ├─ Current plan: Basic ├─ Usage patterns: Heavy data user ├─ Recommended action: Suggest upgraded plan + add-ons └─ Next-best-action: "Present Premium with extras" Agent: "Great! Based on your usage, I have a perfect recommendation..." Copilot displays: ├─ Customer's usage metrics ├─ Recommended plan details ├─ Comparison of savings └─ Available promotions (10% if upgrade today) [Agent presents recommendations] Customer: "The Premium plan looks good, and 10% off sounds great!" Copilot suggests: ├─ Add-on: Premium Support (93% value match) └─ Upsell: Extended warranty (67% match) Result: Upgraded plan + 1 add-on sale Copilot Summary: "Customer upgraded from Basic to Premium plan and added Premium Support. Mentioned promotional discount influenced decision. Total value: $XXX. Customer satisfied." Business Impact: Increased revenue, improved satisfaction ``` --- ## Best Practices ### Knowledge Base Quality - **Accuracy First** - All information must be current and correct - **Comprehensive Coverage** - Address common issues and scenarios - **Clear Language** - Write for quick scanning, not detailed reading - **Proper Organization** - Use tags and metadata effectively - **Regular Updates** - Review and update quarterly minimum - **Validation Process** - Test before publishing ### Agent Copilot Configuration - **Start Simple** - Enable basic features first, add complexity gradually - **Queue-Specific** - Tailor for each queue's unique needs - **Knowledge Curation** - Surface only relevant articles - **Test Thoroughly** - Pilot before full deployment - **Monitor Adoption** - Track agent usage and feedback - **Continuous Refinement** - Improve based on data ### Agent Enablement - **Comprehensive Training** - Thorough explanation of features - **Live Demonstrations** - Show real examples and use cases - **Practice Sessions** - Let agents use Copilot before production - **Clear Benefits** - Help agents understand time-saving value - **Easy Access to Help** - Support for questions and issues - **Celebrate Successes** - Share agent stories and improvements ### Performance Optimization - **Track Metrics Daily** - Monitor AHT, ACW, knowledge usage - **Gather Agent Feedback** - Regular surveys and check-ins - **Analyze Usage Patterns** - Identify which features are most helpful - **Refine Knowledge** - Update articles based on usage data - **Optimize for Your Business** - Tailor features to your priorities - **Benchmark Performance** - Compare pre/post Copilot metrics --- ## Token Consumption Agent Copilot requires AI Experience tokens for operation: - **Consumption**: 40-60 tokens per user per month - **Factors**: Interaction volume, knowledge article access, AI feature usage - **Optimization**: Monitor usage and optimize knowledge articles - **Budgeting**: Plan tokens based on agent count and deployment scope --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What is Agent Copilot? | Real-time AI assistance for agents during customer interactions | | What does it do during calls? | Determines intent, suggests next actions, surfaces knowledge, monitors sentiment | | What about after-call work? | Auto-generates summaries, suggests wrap-up codes, tracks resolution | | What are AI Skills? | Modular AI capabilities (knowledge, next actions, sentiment, compliance) | | How does it behave on-queue? | Continuous monitoring and assistance throughout interaction | | What languages does it support? | 13+ including English, Spanish, French, German, Japanese, Korean, Arabic, Hindi | | How much time does it save? | ~9 minutes per interaction (5 AHT + 2 ACW + 1.5 hold time) | | What's the expected improvement? | AHT reduction 5-10%, ACW reduction 2-3 minutes, CSAT improvement 5-15% | | How is it licensed? | Per-user with 40-60 tokens per user per month | | Does it work omnichannel? | Yes - voice, chat, email all supported | | Can you customize it? | Yes via AI Studio to create custom Guides and Skills | | Where is it configured? | Admin → Contact Center → Agent Assistance → Agent Copilot | | What knowledge does it need? | Quality knowledge base (Workbench or Fabric) | | How long until ROI? | 2-4 weeks to see AHT improvements | | What's most important? | Quality knowledge base and agent training | --- ## Key Takeaways - **Real-Time Assistance** - Continuously monitors interactions providing guidance as they happen - **Intelligent Understanding** - Determines customer intent and emotional state - **Automatic Documentation** - Generates summaries and suggests wrap-up codes - **Knowledge Integration** - Surfaces relevant articles without agent search - **Multi-Channel** - Works on voice, chat, email, and messaging - **Modular Skills** - Combine capabilities for your specific needs - **Proven Results** - 5-minute AHT reduction, 1.5-minute hold time reduction - **Adaptive Behavior** - Responds to conversation flow and customer emotion - **Global Support** - Operates in 13+ languages - **Continuous Learning** - Improves recommendations based on outcomes --- ## Additional Resources ### Official Documentation - About Agent Copilot: help.genesys.cloud/articles/about-genesys-agent-copilot/ - Agent Copilot Configuration: help.genesys.cloud/articles/configure-agent-copilot/ - Agent Copilot Deep Dive: genesys.com/blog/post/genesys-cloud-agent-copilot-deep-dive - Agent Copilot FAQs: help.genesys.cloud/faqs/category/agent-copilot/ ### Support & Training - Genesys University: genesys.com/training - Community Forums: https://community.genesys.com - Technical Support: https://support.genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation **Validated:** Current with January-March 2026 releases **Version:** 1.0 # AI Forecasting (WFM) # Genesys PureCloud AI Forecasting (WFM) Documentation ## Study Notes | Topic | Description | |---|---| | AI Forecasting | Automated, cloud-based forecasting using advanced ML algorithms | | Automatic Best Method (ABM) | Genesys' proprietary ensemble technique selecting optimal algorithm | | Accuracy Focus | Eliminates manual method selection through automation | | Cloud-Based | Automatically updated with latest algorithms and research | | Data Normalization | Handles outliers, missing data, seasonality automatically | | Setup | Simple integration between WFM and Genesys Cloud CX | --- ## Navigation WFM → Forecast → Build Volumes/AHT OR WFM Supervisors → Forecasting (New UI) → Create Forecast --- ## AI Forecasting Overview Genesys Cloud's AI-Powered Forecasting feature is a sophisticated build method that leverages best practices in data science and the industry. It provides users with an easy-button approach to an otherwise complex operation of predicting the workload and service time of agents for contact center planning. The AI-powered forecasting service uses advanced machine learning algorithms to analyze historical data and generate highly accurate forecasts for workforce management, representing an industry-first capability that enables the fastest, most accurate AI-powered forecasting service for better workforce management. ### Key Capabilities - Automated method selection (no manual algorithm choosing required) - Analyzes historical data to determine best forecasting approach - Automatically detects and handles outliers and anomalies - Fills gaps in incomplete historical data - Accounts for seasonality and trends - Continuously evolving - automatically updated with latest algorithms - Cloud-based delivery ensures always-current capability - Supports both volume and AHT (Average Handling Time) forecasting - Works across all contact center channels - Integration with WFM for scheduling and planning ### Performance Improvements Organizations using AI Forecasting typically see: - 15-25% improvement in forecast accuracy over traditional methods - Reduced forecasting time (hours vs. weeks) - Elimination of subjective method selection - Better capacity planning and staffing accuracy - Improved service level achievement - Reduced labor costs through better predictions --- ## Automatic Best Method (ABM) Explained ### What is ABM? Automatic Best Method (ABM) uses AI and ensemble techniques to create highly accurate forecasts for workforce management. Rather than requiring forecasters to manually select which algorithm to use (Expert Average Engine, Universal Modeling Engine, etc.), ABM automatically evaluates multiple algorithms and selects the best one for your specific data. ### How ABM Works ``` Historical Data Input ↓ Data Analysis & Preparation ├─ Identify patterns ├─ Detect seasonality ├─ Find trends └─ Locate anomalies ↓ Evaluate Multiple Algorithms ├─ Expert Average Engine ├─ Universal Modeling Engine ├─ Exponential Smoothing ├─ ARIMA Models ├─ Ensemble Combinations └─ Neural Networks ↓ Machine Learning Selection ├─ Test each algorithm ├─ Compare accuracy metrics ├─ Validate against holdout data ├─ Score performance └─ Select best performer ↓ Generate Forecast ├─ Apply selected algorithm ├─ Apply ensemble weighting ├─ Output predictions └─ Provide confidence intervals ↓ Forecast Delivered ├─ Volume forecast ├─ AHT forecast ├─ Accuracy metrics └─ Ready for scheduling ``` ### ABM vs. Traditional Methods | Aspect | Traditional Methods | Automatic Best Method | |---|---|---| | Method Selection | Manual by forecaster | Automatic via ML | | Algorithm Options | 4-5 choices | 10+ evaluated | | Time Required | Hours to days | Minutes | | Accuracy | Good (70-80%) | Excellent (85-95%+) | | Expertise Required | High (data science knowledge) | Low (point and click) | | Consistency | Variable (person-dependent) | Consistent (algorithmic) | | Updates | Manual recalculation | Automatic weekly | | Optimization | Limited | Full ensemble evaluation | --- ## Edition & Module Requirements | Requirement | Details | |---|---| | Minimum Edition | Genesys Cloud CX 1-4 or Genesys Multicloud CX | | Module | AI-Powered Forecasting add-on (wfmAiPoweredForecasting key) | | WFM Version | WFM 8.5.214 or later | | Setup | OAuth client and Genesys Cloud CX Tier 3 organization | | Integration | Cloud-based connection between WFM and Genesys Cloud | | Internet | WFM servers must have internet access (or proxy) | --- ## Study Notes - AI Forecasting Data Handling | Data Issue | How ABM Handles It | Result | |---|---|---| | Missing Data | Intelligent interpolation and statistical methods | Complete forecasts without gaps | | Outliers | Detects and normalizes extreme values | Realistic forecasts without spikes | | Seasonality | Identifies recurring patterns | Accurate seasonal adjustments | | Trends | Analyzes long-term direction | Projects growth/decline correctly | | Holidays | Adjusts for expected non-working days | Prevents overstaff in holidays | | Spikes | Separates temporary from permanent changes | Distinguishes one-off from real change | | Multiple Channels | Combines data intelligently | Unified forecasts across channels | --- ## How AI Forecasting Improves Accuracy ### Traditional Forecasting Challenges ``` Manual Method Selection Process: 1. Forecaster reviews 6 months of data 2. Attempts Expert Average Engine └─ Accuracy: 72% 3. Tries Universal Modeling Engine └─ Accuracy: 78% 4. Tests Exponential Smoothing └─ Accuracy: 75% 5. Manual decision: Use Universal Modeling (78%) 6. Generate forecast Problems: ├─ Time-consuming (hours) ├─ Subjective decision-making ├─ Might not find best option ├─ Requires expertise └─ Accuracy: ~78% ``` ### AI Forecasting Approach ``` Automatic Best Method Process: 1. Historical data loaded into Genesys Cloud 2. ABM evaluates 10+ algorithms in parallel ├─ Expert Average Engine: 72% ├─ Universal Modeling: 78% ├─ Exponential Smoothing: 75% ├─ ARIMA: 81% ├─ Ensemble Combination A: 87% ├─ Ensemble Combination B: 89% ├─ Neural Networks: 86% ├─ Bayesian Approach: 84% └─ [Additional algorithms...] 3. ML system selects: Ensemble Combination B (89%) 4. Generate forecast Benefits: ├─ Automatic (minutes) ├─ Objective evaluation ├─ Finds optimal option ├─ No expertise required ├─ Higher accuracy: ~89% ``` ### Accuracy Improvements by Scenario **High-Volatility Queue** ``` Traditional Method: 68% accuracy AI Forecasting: 84% accuracy Improvement: +16 points (24% better) Business Impact: Better staffing decisions, fewer shortages ``` **Seasonal Business** ``` Traditional Method: 74% accuracy AI Forecasting: 91% accuracy Improvement: +17 points (23% better) Business Impact: Optimized staffing for season changes ``` **Multi-Channel Center** ``` Traditional Method: 71% accuracy AI Forecasting: 87% accuracy Improvement: +16 points (23% better) Business Impact: Unified forecasting across channels ``` ### Why ABM is More Accurate 1. **Exhaustive Algorithm Evaluation** - Tests multiple approaches, not just a few 2. **Ensemble Methods** - Combines best algorithms for optimal accuracy 3. **Continuous Optimization** - Cloud service updates automatically 4. **Machine Learning** - Adapts to your specific data patterns 5. **Latest Research** - Incorporates cutting-edge forecasting algorithms 6. **No Human Bias** - Objective selection based on data, not opinion 7. **Validation** - Tests against holdout data before deployment --- ## Implementation Guide ### Step 1: Prerequisites 1. Verify Genesys Cloud CX organization created (Tier 3) 2. Confirm WFM version 8.5.214 or later 3. Obtain wfmAiPoweredForecasting product key 4. Verify internet connectivity for WFM servers 5. Create OAuth client in Genesys Cloud (Client Credentials grant) 6. Gather required credentials and configuration info ### Step 2: Configure WFM for AI Forecasting 1. Open Genesys Administrator 2. Navigate to WFM Server Applications 3. Set authentication provider to WFM 4. Configure purecloud section: ``` [auth] provider = wfm [PureCloud] purecloud.client_id = purecloud.client_secret = purecloud.upload_uri = https://apps.mypurecloud.com purecloud.region = ``` 5. Save configuration 6. Restart WFM Server ### Step 3: Enable in WFM UI 1. Log into WFM Supervisors interface 2. Navigate to Forecasting 3. In New UI settings, enable "Use Latest Forecast UI" 4. Verify AI Forecasting option appears 5. Test data connectivity ### Step 4: Create First Forecast 1. Select queue/skill for forecasting 2. Review historical data (at least 90 days recommended) 3. Choose forecast type: - Volume forecast (interaction count) - AHT forecast (handling time) - Combined forecast (both) 4. Select time period to forecast 5. Choose "AI-Powered Forecasting" method 6. System automatically: - Analyzes data - Evaluates algorithms - Selects best method - Generates forecast 7. Review results and accuracy metrics 8. Save and publish forecast ### Step 5: Validation & Testing 1. Compare AI forecast accuracy to historical actual 2. Review outlier handling 3. Validate seasonality adjustments 4. Test with upcoming actual data 5. Monitor variance between forecast and actual 6. Refine as needed ### Step 6: Production Use 1. Generate forecasts on regular cadence (weekly recommended) 2. Use for schedule building in WFM 3. Monitor forecast accuracy continuously 4. Adjust refresh frequency if needed 5. Establish process for anomalies --- ## Real-World Implementation Scenario ### Mid-Market Contact Center ``` Organization: Financial Services Company Current: 250 agents across 4 skill groups Challenge: Manual forecasting taking 16 hours/week Accuracy: ~74% (missing peaks/valleys) Implementation: Week 1: Setup ├─ Created Genesys Cloud Tier 3 org ├─ Configured OAuth client ├─ Updated WFM to 8.5.214 └─ Integrated WFM with Cloud Week 2-3: First Forecasts ├─ Migrated 6 months of historical data ├─ Generated AI forecasts for each skill group ├─ Compared to previous manual method └─ Accuracy improved from 74% to 87% Results in First 30 Days: ├─ Forecast time: 16 hours → 2 hours (87% reduction) ├─ Accuracy improvement: +13 points (74% → 87%) ├─ Better staffing alignment ├─ Fewer service level misses ├─ Reduced overtime due to better predictions └─ Cost savings: ~$4,000/month Year 1 Impact: ├─ Consistent 85-88% forecast accuracy ├─ Service levels improved 5-8 points ├─ Labor costs down 3-5% ├─ Forecaster time freed for analysis └─ ROI: Payback in 3 months ``` --- ## AI Forecasting vs. Traditional Methods ### Method Comparison **Expert Average Engine** - Best for: Stable, consistent patterns - Accuracy: ~75% - Time: Manual selection (~1-2 hours) - Pros: Fast if chosen correctly - Cons: Misses complex patterns **Universal Modeling Engine** - Best for: Moderate complexity - Accuracy: ~78% - Time: Manual selection (~2-3 hours) - Pros: Handles trends well - Cons: Not optimal for all scenarios **Template-Based** - Best for: Similar queues - Accuracy: ~70% - Time: Minutes - Pros: Quick - Cons: Limited accuracy **Automatic Best Method (AI)** - Best for: All scenarios - Accuracy: 85-92% - Time: Minutes (automatic) - Pros: Always optimal, no expertise needed - Cons: Requires cloud integration --- ## Best Practices ### Data Quality - **Historical Data** - Maintain at least 90 days, ideally 180+ days - **Data Accuracy** - Ensure interaction counts and AHT are correct - **Clean Data** - Remove obvious data entry errors - **Consistent Definitions** - Define queues/skills consistently - **Regular Validation** - Verify data quality before forecasting ### Forecasting Process - **Regular Cadence** - Generate forecasts weekly at minimum - **Validation** - Compare forecasts to actuals and adjust - **Anomaly Handling** - Identify unusual events and exclude if needed - **Multiple Methods** - Consider different forecast scenarios - **Document Changes** - Track what changed between forecasts ### Integration with Scheduling - **Use Forecasts Immediately** - Apply to schedule building - **Schedule Alignment** - Ensure schedules match forecast expectations - **Adherence Tracking** - Monitor agents vs. schedule accuracy - **Feedback Loop** - Use actual data to improve next forecast - **Continuous Improvement** - Refine process over time ### Organizational Adoption - **Stakeholder Buy-In** - Get forecasting team support - **Clear Communication** - Explain benefits and changes - **Training** - Educate on new AI method - **Quick Wins** - Show improvements early - **Continuous Monitoring** - Track and share successes --- ## Common Forecasting Scenarios ### Scenario 1: Intraday Forecasting ``` Monday 8 AM Forecast for Monday 1 PM - 6 PM Historical Pattern: ├─ Lunch hour (12-1 PM): 50% volume drop ├─ Afternoon (1-3 PM): Recovery to 90% normal ├─ Late afternoon (3-6 PM): Peak (110% normal) └─ Volume trend: Growing 2% week-over-week AI Forecast Output: ├─ 1 PM: 45 inbound calls (50% of 90) ├─ 2 PM: 82 inbound calls (91% of 90) ├─ 3 PM: 95 inbound calls (105% of 90) ├─ 4 PM: 100 inbound calls (111% of 90) ├─ 5 PM: 98 inbound calls (109% of 90) └─ 6 PM: 85 inbound calls (94% of 90) AHT Pattern: ├─ Lunch hour: +8% (simpler issues) ├─ Afternoon: +2% (baseline) └─ Evening: +5% (more complex) Result: Accurate staffing for afternoon peak ``` ### Scenario 2: Weekly Forecast (Monday-Friday) ``` Week of March 10-14 Pattern: Business process deadline Friday ├─ Monday: 90% normal volume ├─ Tuesday: 85% normal volume (preparation starts) ├─ Wednesday: 110% normal volume (deadline week) ├─ Thursday: 115% normal volume (peak) └─ Friday: 120% normal volume (final deadline) AI Detects: Weekly recurring pattern Adjusts: Volume forecast accordingly Result: Optimal staffing for each day ``` ### Scenario 3: Holiday Period ``` Christmas Week Forecast Pattern Recognition: ├─ December 23: Normal (Monday before holiday) ├─ December 24: 140% volume (last-minute issues) ├─ December 25: Closed (holiday) ├─ December 26: 160% volume (backup) └─ December 27-31: 120% (customers want help before year-end) AI Handles: Automatically adjusts for holiday Result: Accurate staffing despite disruption ``` --- ## Monitoring & Optimization ### Forecast Accuracy Tracking - **Daily** - Compare forecast to actual volume/AHT - **Weekly** - Calculate accuracy percentage - **Monthly** - Analyze trends and patterns - **Quarterly** - Review overall accuracy and improvements - **Annually** - Strategic assessment of forecasting effectiveness ### Accuracy Metrics | Metric | Formula | Target | |---|---|---| | Mean Absolute Percentage Error (MAPE) | Avg \|(Actual - Forecast) / Actual\| | <10% | | Mean Bias | Avg(Actual - Forecast) | ±3% | | Peak Accuracy | Accuracy during peak times | >85% | | Valley Accuracy | Accuracy during low-volume times | >80% | ### Continuous Improvement - **Review Forecast-Actual Variance** - Identify patterns of over/under forecasting - **Adjust for New Patterns** - Update forecasts as business changes - **Incorporate Feedback** - Use scheduling/staffing feedback - **Test Scenarios** - Model impact of business changes - **Optimize Parameters** - Fine-tune forecasting settings --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What is AI Forecasting? | Cloud-based automated forecasting using ML algorithms | | What's ABM? | Automatic Best Method - automatically selects optimal algorithm | | How is it different from traditional? | No manual method selection; evaluates 10+ algorithms automatically | | What accuracy improvement? | Typically 15-25% better than traditional methods (85-92% vs 70-80%) | | How does it handle anomalies? | Automatically detects and normalizes outliers and missing data | | What data is needed? | 90+ days minimum, ideally 180+ days of historical data | | Is it cloud-only? | Yes, requires Genesys Cloud CX Tier 3 organization | | How long to setup? | 1-2 weeks for initial configuration and testing | | What WFM version? | WFM 8.5.214 or later | | Does it handle seasonality? | Yes, automatically detects and accounts for seasonal patterns | | Can it forecast multiple channels? | Yes, combines data intelligently across all channels | | How often to refresh? | Weekly minimum, can be daily for high-variance queues | | What's the ROI? | Typically 3-4 month payback through better staffing | | Can it predict anomalies? | Detects anomalies but requires manual handling of known events | | What metrics are tracked? | Volume, AHT, accuracy, forecast vs. actual variance | --- ## Key Takeaways - **Automatic Selection** - No manual algorithm choosing; AI finds optimal method - **High Accuracy** - 85-92% typical accuracy vs. 70-80% traditional - **Industry-Leading** - Fastest, most accurate AI-powered forecasting available - **Easy to Use** - Simple point-and-click interface, no expertise needed - **Intelligent Data Handling** - Automatically manages outliers, seasonality, missing data - **Cloud-Based** - Always updated with latest algorithms - **Faster Process** - Hours/weeks reduced to minutes - **Ensemble Methods** - Combines multiple algorithms for optimal results - **Continuous Learning** - Improves recommendations over time - **Proven ROI** - 3-4 month payback through better staffing --- ## Additional Resources ### Official Documentation - Forecasting Overview: all.docs.genesys.com/PEC-WFM/GFrcstg - ABM Overview: help.mypurecloud.com/articles/automatic-best-method-forecast-method-overview/ - AI-Powered Forecasting Setup: docs.genesys.com/Documentation/WM/latest/Admin/AIPwrdFrcst - Multicloud CX AI Forecasting: all.docs.genesys.com/PEC-WFM/PECAIPwrd ### Support & Training - Genesys University: genesys.com/training - Community Forums: https://community.genesys.com - Technical Support: https://support.genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation **Validated:** Current with January-March 2026 releases **Version:** 1.0 # Supervisor Copilot # Genesys PureCloud Supervisor Copilot Documentation ## Study Notes | Topic | Description | |---|---| | Supervisor Copilot | AI-powered assistant providing real-time insights for supervisors | | Virtual Supervisor | Automates evaluation scoring using AI for 100% of interactions | | AI Scoring | Automated performance evaluations with transparency and fairness | | Analytics Explorer | AI Skill providing historical and real-time data insights | | Quality Management | Automates routine tasks, focuses supervisors on coaching | | Time Savings | 40% reduction in quality evaluation time, 25% in multilingual work | --- ## Navigation Admin → Quality Management → Virtual Supervisor OR Quality → Evaluation Dashboard → AI Scoring Settings --- ## Supervisor Copilot Overview Supervisor Copilot is an AI-powered toolkit designed to enhance supervisor productivity and decision-making in the contact center. It acts as a sidekick for managers, providing prescriptive support for quality assurance, compliance and coaching. Powered by generative AI, it automatically summarizes interactions, allowing supervisors to quickly review and make informed decisions. Virtual Supervisor and Supervisor Copilot are AI-powered tools designed to support and enhance supervisor workflows. These features combine powerful capabilities (AI scoring, summary, insights, and translate), to deliver a smarter, more efficient way to monitor performance, gain clarity from conversations, and act quickly on key information. ### Key Capabilities - AI Scoring for automated performance evaluations of 100% of interactions - AI Summary & Insights capturing full conversation context - AI Translate converting transcripts into 70+ languages - Automated interaction summaries highlighting key moments - Reason for contact, resolution, action items, sentiment drivers - Advanced quality and conversational intelligence - Compliance monitoring across interactions - Coaching opportunity identification - Real-time performance insights - Integration with Agent Copilot data ### Performance Improvements Organizations using Supervisor Copilot report: - 40% reduction in quality evaluation time - 25% reduction in multilingual evaluation time - 38% decrease in quality management administrative costs - Better coaching effectiveness through data-driven insights - Improved compliance and consistency - Enhanced agent performance and satisfaction --- ## Edition & Module Requirements | Requirement | Details | |---|---| | Minimum Edition | Genesys Cloud CX 1-4 (CX 3-4 recommended) | | Module | Virtual Supervisor or Supervisor Copilot add-on | | License Type | Required for quality scoring and analytics | | AI Tokens | Consumption based on evaluation volume | | Features | AI Scoring, AI Translate, AI Summary & Insights | --- ## Study Notes - Supervisor Copilot Components | Component | Function | Benefit | |---|---|---| | AI Scoring | Automated evaluation of interactions | 100% coverage, consistency, reduced manual work | | AI Translate | Multi-language transcript conversion | Faster multilingual review, 25% time reduction | | AI Summary & Insights | Key points extraction and analysis | Quick understanding, identified coaching needs | | Reason for Contact | Automatic issue categorization | Pattern identification, trend analysis | | Resolution Status | Whether issue was resolved | Outcome tracking, quality assessment | | Action Items | Follow-up tasks identified | Process improvement, compliance tracking | | Sentiment Drivers | What caused customer emotion | Root cause analysis, service improvement | | Analytics Explorer | Data visualization and trends | Real-time insights, strategic visibility | --- ## Virtual Supervisor & AI Scoring ### What is Virtual Supervisor? Virtual Supervisor enhances Quality Management by leveraging AI to automate interaction evaluations. It automatically scores interactions and delivers actionable insights, highlighting areas for improvement and explaining the rationale behind each assessment. By reducing the need for manual reviews, Virtual Supervisor empowers supervisors to more effectively support agents and focus on coaching and performance development. ### How AI Scoring Works ``` Interaction Completed ↓ Virtual Supervisor Receives Transcript ├─ Recording ├─ Transcript ├─ Agent info └─ Customer info ↓ AI Analysis Engine Evaluates ├─ Compliance requirements ├─ Quality standards ├─ Best practices ├─ Behavioral criteria └─ Customer outcomes ↓ Scoring Process ├─ Question 1: Did agent greet properly? │ └─ AI Analysis: YES - "Agent used proper greeting with company name" │ ├─ Question 2: Compliance disclosure given? │ └─ AI Analysis: YES - "Agent stated privacy policy per minute 2:15" │ ├─ Question 3: Issue resolved? │ └─ AI Analysis: YES - "Customer confirmed satisfaction at end" │ ├─ Question 4: Proper tone used? │ └─ AI Analysis: PARTIAL - "Professional but slightly rushed in places" │ └─ Question 5: Call handled efficiently? └─ AI Analysis: YES - "Average handle time 6.2 min vs queue avg 6.8" ↓ Score Generated ├─ Greeting: 10/10 ├─ Compliance: 10/10 ├─ Resolution: 10/10 ├─ Tone: 7/10 ├─ Efficiency: 9/10 └─ Overall Score: 92/100 ↓ Justifications Provided ├─ Strengths: Compliance excellent, efficient handling ├─ Improvement Areas: Pacing during explanations └─ Coaching Recommendations: Practice clear, methodical explanations ↓ Delivered to Supervisor ├─ Score with justifications ├─ Transcript highlighting ├─ Coaching suggestions └─ Comparison to standards ``` ### AI Scoring Features **Automated Scoring** - Scores 100% of interactions (not just sample) - Consistent application of standards - Eliminates evaluator bias - 24/7 operation (no human scheduling) **Transparency & Fairness** - AI-generated justifications for each score - Shows reasoning behind assessments - Highlights both strengths and improvements - Supervisors can review and adjust before final score - Promotes fair and consistent evaluation **Multi-Question Coverage** - Behavioral criteria (tone, empathy, etc.) - Compliance requirements (disclosures, adherence) - Process adherence (proper steps followed) - Customer outcomes (satisfaction, resolution) - Efficiency metrics (handle time, efforts) **Quality Control** - Supervisors review and approve/adjust scores - Two-step process ensures accuracy - Learn from supervisor overrides - Continuous model improvement --- ## Analytics Explorer (AI Skill) Analytics Explorer is the first AI Skill that provides historical and real-time data to help supervisors lower time-to-insight and access performance trends without complex dashboards. ### What is Analytics Explorer? Analytics Explorer uses natural language processing to let supervisors ask questions about metrics and trends in plain language, rather than navigating complex dashboards. It provides: - Historical performance data - Real-time metrics - Trend analysis - Comparative insights - Anomaly detection - Predictive indicators ### How Analytics Explorer Works ``` Supervisor Question: "Which agent had the highest AHT this week?" Natural Language Processing ├─ Identify: Query type (agent comparison) ├─ Extract: Metric (AHT) ├─ Parse: Timeframe (this week) └─ Understand: Ranking (highest) ↓ Data Query & Analysis ├─ Retrieve: This week's AHT data for all agents ├─ Calculate: Average per agent ├─ Rank: By highest to lowest └─ Context: Compare to queue average ↓ Response Generated "Agent James had the highest AHT this week at 8.3 minutes, which is 15% above queue average of 7.2 minutes. This is typical for his skill set but elevated compared to his personal average of 7.8." ↓ Additional Insights ├─ Trend: Increasing vs past 4 weeks ├─ Context: Why (queue complexity, etc.) └─ Recommendation: Review calls, provide coaching ``` ### Common Analytics Explorer Queries **Performance Trends** - "What's Sarah's average handle time trend?" - "Which queues had lowest CSAT this month?" - "Show me top performers on calls answered" **Comparative Analysis** - "How does Team A compare to Team B on quality?" - "Which channels have lowest first-contact resolution?" - "What's the difference in AHT between morning and afternoon?" **Anomaly Detection** - "Why is call volume so high today?" - "Which interactions had unusual sentiment?" - "Show me abandonment spike causes" **Predictive Insights** - "What's the forecast for queue volume Friday?" - "Which agents might need coaching based on trends?" - "What's the predicted impact of staffing changes?" --- ## Supervisor Workflows with Copilot ### Quality Management Workflow ``` Traditional vs. AI-Powered Approach: TRADITIONAL (16 hours/supervisor/week): ├─ Monday: Randomly select 25 interactions ├─ Tuesday: Listen to calls (2 hours) ├─ Wednesday: Score evaluations (2 hours) ├─ Thursday: Compile feedback (1 hour) ├─ Friday: Meet with agents individually (6 hours) └─ Continuous: Handle urgent issues AI-POWERED (10 hours/supervisor/week): ├─ Morning: Review AI Scoring of 100% interactions │ └─ 1 hour to review 150+ scored interactions ├─ Identify: Focus areas and top performers │ └─ 30 min (automated pattern detection) ├─ Deep Dives: Review specific interactions │ └─ 2 hours (only complex or critical ones) ├─ Coaching: Deliver targeted feedback │ └─ 4 hours (focused on high-impact areas) ├─ Analytics: Trend analysis and planning │ └─ 1.5 hours (data-driven decisions) └─ Strategic: Process improvement and planning └─ 1 hour (freed-up supervisory time) Time Savings: 6 hours/week per supervisor = $12,000/year per supervisor Quality Improvement: Data-driven coaching, 100% coverage vs. sampling ``` ### Coaching Workflow with AI Insights ``` Step 1: AI Scoring Identifies Issues ├─ Virtual Supervisor scores 100 interactions ├─ 15 need coaching on "tone and empathy" ├─ 8 have compliance gaps └─ 3 show efficiency concerns Step 2: Supervisor Reviews AI Justifications ├─ Reads AI explanations for each score ├─ Understands root causes ├─ Sees agent strengths highlighted └─ Identifies coaching themes Step 3: Supervisor Selects Coaching Targets ├─ Priority 1: Compliance gaps (critical) ├─ Priority 2: Tone issues (systemic) ├─ Priority 3: Individual efficiency (targeted) └─ Recognition: Top performers (5 agents) Step 4: Prepare Coaching Sessions ├─ Pull specific interaction examples ├─ Write coaching notes with AI suggestions ├─ Plan 1-on-1 meetings └─ Prepare recognition for strong performers Step 5: Conduct Coaching ├─ Share AI insights with agents ├─ Discuss specific examples ├─ Explain improvements with data ├─ Create development plans └─ Recognize strengths and efforts Step 6: Follow-Up ├─ Monitor next evaluations ├─ Use AI Scoring to track improvement ├─ Celebrate progress with agents └─ Adjust coaching if needed ``` --- ## Real-Flow Scenarios ### Scenario 1: Automated Quality Evaluation ``` Monday Morning - Quality Review: Supervisor logs in at 9 AM AI Scoring Summary Shows: ├─ 147 interactions scored over weekend ├─ Average quality score: 87/100 ├─ 3 interactions scored below 70 (below standard) ├─ 12 interactions with compliance tags ├─ 8 agents exceeded their personal average └─ 2 high-risk interactions flagged Supervisor Action (30 minutes): 1. Reviews 3 low-scoring interactions ├─ Reads AI justifications ├─ Listens to specific moments └─ Identifies coaching points 2. Flags 12 compliance interactions ├─ Shares with quality team ├─ Schedules compliance training └─ Creates preventive alerts 3. Celebrates high performers ├─ Sends recognition messages ├─ Shares with management └─ Motivates team Result: What would take 4-5 hours manually is done in 30 minutes Quality: More objective, data-driven, fair evaluations Coaching: More targeted, based on data insights ``` ### Scenario 2: Multilingual Evaluation ``` Wednesday - International Team Review: Supervisor manages team across 3 languages: ├─ English speakers (50%) ├─ Spanish speakers (35%) └─ French speakers (15%) Traditional Approach: ├─ Must hire bilingual evaluators ├─ Or use translation services (slow, expensive) ├─ Or only evaluate in English (unfair) └─ Result: Inconsistent quality evaluation AI-Powered Approach: ├─ Virtual Supervisor AI Translates all transcripts ├─ All agents evaluated in supervisor's language ├─ Consistent standards across all languages ├─ Takes 25% of the time vs traditional methods └─ Result: Fair, consistent, efficient Supervisor Benefits: ├─ Evaluates all agents fairly ├─ No language barrier ├─ Consistent quality standards ├─ Time saved: 75% reduction in multilingual work └─ No translation costs ``` ### Scenario 3: Compliance Monitoring ``` Friday End-of-Day - Compliance Check: Supervisor needs to verify compliance for audit: AI Scoring automatically: ├─ Reviewed 500 interactions this week ├─ Tagged required disclosures (100% coverage) ├─ Flagged missing compliance statements ├─ Noted proper documentation └─ Generated compliance report Supervisor Action (30 minutes): ├─ Reviews AI-generated compliance report ├─ Drills into 3 flagged interactions ├─ Notes pattern (new agents missing disclosure) ├─ Schedules training session └─ Reports 98% compliance to management Traditional Manual Approach: ├─ Would require sampling (maybe 10%) ├─ Time: 8+ hours of listening ├─ Coverage: ~10% of interactions ├─ Risk: Might miss issues └─ Inefficient for audit AI Result: 100% coverage, done in 30 min, comprehensive audit-ready report ``` --- ## Best Practices ### Quality Management - **Leverage AI Scoring** - Use 100% scoring to identify true patterns - **Review Justifications** - Understand AI reasoning before coaching - **Focus on High-Impact Areas** - Use data to prioritize coaching - **Consistent Standards** - AI ensures consistent application - **Regular Monitoring** - Use automated scoring for continuous improvement ### Agent Coaching - **Share Data** - Show agents the AI scoring and justifications - **Be Fair** - Explain that all agents are scored equally - **Focus on Development** - Use AI insights for targeted coaching - **Celebrate Strengths** - Recognize what agents do well - **Track Progress** - Monitor improvements with AI scores ### Compliance & Governance - **Regular Audits** - Use AI Scoring for continuous compliance tracking - **Investigate Flags** - Review AI-flagged compliance issues - **Training Reinforcement** - Provide coaching on compliance gaps - **Documentation** - Keep records for audit purposes - **Preventive Measures** - Address systemic issues proactively ### Analytics & Insights - **Use Analytics Explorer** - Ask natural language questions about data - **Trend Analysis** - Identify patterns and anomalies - **Comparative Insights** - Understand relative performance - **Forecasting** - Use predictive indicators for planning - **Data-Driven Decisions** - Base coaching and scheduling on evidence --- ## Implementation Guide ### Step 1: Enable Supervisor Copilot 1. Navigate to Admin → Quality Management 2. Enable "Supervisor Copilot" features 3. Configure AI Scoring permissions 4. Set up Virtual Supervisor access 5. Enable Analytics Explorer AI Skill ### Step 2: Configure AI Scoring 1. Define quality evaluation form 2. Map questions to AI scoring categories 3. Set scoring standards and thresholds 4. Configure supervisor review process 5. Establish approval workflow ### Step 3: Train Supervisors 1. Explain AI Scoring capabilities 2. Show how to review AI justifications 3. Practice using Analytics Explorer 4. Establish new coaching workflow 5. Share best practices ### Step 4: Establish Workflow 1. Define daily/weekly quality reviews 2. Establish coaching process 3. Create compliance monitoring procedures 4. Set up reporting and analysis 5. Establish escalation process ### Step 5: Monitor & Optimize 1. Track adoption metrics 2. Gather supervisor feedback 3. Review AI Scoring accuracy 4. Optimize based on learnings 5. Refine training as needed --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What is Supervisor Copilot? | AI-powered assistant providing insights and automation for supervisors | | What does Virtual Supervisor do? | Automates evaluation scoring of 100% of interactions with justifications | | How much time is saved? | 40% reduction in quality evaluation time, 25% in multilingual work | | What is AI Scoring? | Automated evaluation using AI with transparency and fairness | | How does fairness work? | AI-generated justifications explain every score, supervisors can adjust | | What about multilingual? | AI Translate converts transcripts to 70+ languages automatically | | What's Analytics Explorer? | AI Skill providing natural language access to metrics and trends | | Can supervisors override AI? | Yes, review and adjust scores before finalizing | | What's included in summary? | Reason for contact, resolution, action items, sentiment drivers | | How accurate is AI Scoring? | Uses latest LLMs; continuously improving accuracy | | Can it handle compliance? | Yes, flags compliance issues and monitors requirements | | What's the ROI? | 40% time savings + better quality through consistency | | How long to implement? | 2-4 weeks for setup and training | | What edition needed? | CX 3-4 recommended (CX 1-2 possible with add-on) | | Does it work with Agent Copilot? | Yes, integrates seamlessly with Agent Copilot data | --- ## Key Takeaways - **Automated Scoring** - Scores 100% of interactions vs. sampling - **Transparency** - AI justifications explain every score - **Fairness** - Consistent, objective evaluation across all agents - **Significant Time Savings** - 40% reduction in quality evaluation time - **Better Coaching** - Data-driven insights enable targeted development - **Multilingual Support** - AI handles 70+ languages automatically - **Compliance** - Automated monitoring and documentation - **Analytics Ready** - Natural language access to all metrics - **Integration** - Works seamlessly with Agent Copilot - **Continuous Improvement** - Learning from interactions and feedback --- ## Additional Resources ### Official Documentation - Virtual Supervisor & Copilot: help.mypurecloud.com/articles/about-virtual-supervisor-and-supervisor-copilot/ - AI Scoring: help.genesys.cloud/articles/about-virtual-supervisor/ - Supervisor Copilot Overview: help.genesys.cloud/articles/about-supervisor-copilot/ - Quality Management AI: help.genesys.cloud/articles/ai-driven-quality-management/ ### Support & Training - Genesys University: genesys.com/training - Community Forums: https://community.genesys.com - Technical Support: https://support.genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation **Validated:** Current with January-March 2026 releases **Version:** 1.0 # AI Tokens & Pricing # Genesys PureCloud AI Tokens & Pricing Documentation ## Study Notes | Topic | Description | |---|---| | Token Model | Consumption-based pricing for AI features | | Monthly Allocation | 250 tokens for named users, 350 for concurrent users | | Pricing | ~$1.00 per additional token beyond allocation | | Consumption Tracking | Per-feature metering with detailed reporting | | Fair Use Policy | Most customers covered by included allocation | | Billing | Monthly renewal with no carryover | --- ## Navigation Admin → Billing & Subscriptions → AI Experience Tokens OR Reports → Usage → AI Token Consumption --- ## AI Tokens Overview Genesys Cloud implements a comprehensive token-based pricing model for AI features across CX 1-4 license tiers. The platform offers a suite of AI capabilities including Agent Copilot, AI Scoring, Translation, Virtual Agent services, predictive routing, and analytics with flexible consumption-based pricing. Genesys Cloud AI Experience tokens help you monitor and manage feature consumption and offer flexibility for changing business needs. Tokenization in AI is a way to track AI engagement in real time by allocating fixed units of measurement to usage costs. This allows organizations to pay only for what they use, providing a scalable, cost-efficient way to integrate AI into operations. ### Default Token Allocations - **Named User Model**: 250 tokens per user per month - **Concurrent User Model**: 350 tokens per month - **Organization Level**: Shared token pool across all users - **Renewal**: Monthly renewal with no carryover - **Additional Tokens**: Available at ~$1.00 per token (pricing varies by currency) --- ## Edition & Module Requirements | Requirement | Details | |---|---| | Minimum Edition | All Genesys Cloud CX tiers include base tokens | | Token Access | All CX 1-4 organizations receive default allocation | | Ordering | New customers have access by default | | Additional Tokens | Purchase through AppFoundry or Genesys representative | | Billing | Included on monthly invoice with consumption tracking | --- ## Study Notes - Token Consumption by Feature | Feature | Unit | Tokens Required | Use Case | |---|---|---|---| | Agent Copilot | Per user | 40-60/month | Real-time agent assistance | | AI Scoring | Per evaluation | 1 token per 20 evals | Quality management | | AI Translate | Per minute | 1 token per 100 min | Multilingual transcripts | | Voice Bots | Per minute | 1 token per 17 min | IVR automation | | Digital Bots | Per session | 1 token per 51 sess | Chat/messaging automation | | Messaging | Per message | 1 token per 400 msg | Direct messaging channels | | Social | Per post | 1 token per 400 post | Social media listening | | Virtual Agent | Per session | 2 tokens per session | Autonomous conversation | | Predictive Routing | Per interaction | Included | Agent routing optimization | | Analytics | Per user | 30-45/month | Analytics access | --- ## Token Consumption Details ### Agent Copilot - **Consumption**: 40-60 tokens per user per month - **Factors Affecting**: - Interaction volume (more interactions = more tokens) - Knowledge article access frequency - AI feature usage intensity (summaries, wrap-up code suggestions) - Number of agents using Copilot - **Example**: 100 agents with heavy Agent Copilot use = 4,000-6,000 tokens/month - **Recommendation**: Budget 50 tokens per user as safe estimate ### AI Scoring (Virtual Supervisor) - **Consumption**: 1 token per 20 evaluations - **Calculation**: - 100 interactions/day × 20 working days = 2,000/month - 2,000 ÷ 20 = 100 tokens/month - **Example**: Medium contact center (400 interactions/day) ≈ 400 tokens/month - **Optimization**: Batch evaluations when possible ### AI Translate - **Consumption**: 1 token per 100 minutes of transcript - **Use Case**: Converting multilingual transcripts - **Example**: 10 hours of multilingual review/month = ~6 tokens - **Application**: Quality management, supervisor reviews ### Voice Bots - **Consumption**: 1 token per 17 minutes - **Calculation**: - 1,000 bot interactions × 5 min avg = 5,000 minutes - 5,000 ÷ 17 = ~294 tokens/month - **Example**: Heavy bot deployment = 300-500 tokens/month - **Optimization**: Improve bot deflection rate to reduce minutes ### Digital Bots - **Consumption**: 1 token per 51 sessions - **Calculation**: - 1,000 bot sessions/month = 1,000 ÷ 51 = ~20 tokens - **Example**: Moderate chat bot use = 50-100 tokens/month - **Optimization**: End sessions efficiently to reduce token use ### Messaging Channels - **Consumption**: 1 token per 400 messages - **Channels**: Facebook, Instagram, WhatsApp, X (Twitter), Apple Business Chat - **Calculation**: - 10,000 messages/month = 10,000 ÷ 400 = 25 tokens - **Example**: Multi-channel deployment = 100-500 tokens/month depending on volume - **Optimization**: High-volume channels provide better token efficiency ### Social Media Listening & Responses - **Consumption**: 1 token per 400 posts - **Application**: Social listening and outbound responses - **Calculation**: - 2,000 social posts/month = 2,000 ÷ 400 = 5 tokens - **Example**: Active social presence = 50-200 tokens/month - **Optimization**: Focus on highest-value channels ### Virtual Agent Sessions - **Consumption**: 2 tokens per session - **Session Definition**: Single customer interaction on any channel - **Calculation**: - 100 virtual agent sessions/day × 20 days = 2,000 sessions - 2,000 × 2 = 4,000 tokens/month - **Example**: High-volume automation = 2,000-5,000+ tokens/month - **Optimization**: Improve automation rates to use fewer sessions ### Predictive Routing - **Consumption**: Included (no token cost) - **Benefit**: Available at no additional cost - **Note**: Foundational AI feature included with platform ### Speech & Text Analytics - **Consumption**: 30-45 tokens per user per month - **Factors**: Licensing type (supervisory vs. analyst) - **Typical Range**: 500-2,000 tokens/month depending on user count ### Predictive Engagement - **Consumption**: No token charges - **Note**: Included AI capability without token consumption --- ## Token Allocation Examples ### Small Contact Center (50 agents) ``` Configuration: ├─ Agent Copilot: 50 agents × 50 tokens = 2,500 ├─ AI Scoring: ~200 interactions/day │ └─ 200 × 20 days ÷ 20 = 200 tokens ├─ Messaging: ~500 messages/month │ └─ 500 ÷ 400 = 2 tokens └─ Virtual Agent: ~50 sessions/day └─ 50 × 20 × 2 = 2,000 tokens Total Monthly: ~4,700 tokens Included: 250 × 50 = 12,500 tokens Overage: None (well under allocation) Cost: $0 overage ``` ### Mid-Market Center (200 agents) ``` Configuration: ├─ Agent Copilot: 200 agents × 55 tokens = 11,000 ├─ AI Scoring: ~500 interactions/day │ └─ 500 × 20 ÷ 20 = 500 tokens ├─ AI Translate: ~5 hours/month │ └─ 300 min ÷ 100 = 3 tokens ├─ Messaging: ~5,000 messages/month │ └─ 5,000 ÷ 400 = 13 tokens ├─ Voice Bots: ~1,000 interactions × 5 min │ └─ 5,000 ÷ 17 = 294 tokens └─ Virtual Agent: ~200 sessions/day └─ 200 × 20 × 2 = 8,000 tokens Total Monthly: ~19,810 tokens Included: 250 × 200 = 50,000 tokens Overage: None Cost: $0 overage ``` ### Enterprise Center (1,000 agents) ``` Configuration: ├─ Agent Copilot: 1,000 agents × 50 tokens = 50,000 ├─ AI Scoring: ~2,000 interactions/day │ └─ 2,000 × 20 ÷ 20 = 2,000 tokens ├─ AI Translate: ~50 hours/month │ └─ 3,000 min ÷ 100 = 30 tokens ├─ Messaging: ~50,000 messages/month │ └─ 50,000 ÷ 400 = 125 tokens ├─ Voice Bots: ~5,000 interactions × 4 min │ └─ 20,000 ÷ 17 = 1,176 tokens ├─ Digital Bots: ~10,000 sessions/month │ └─ 10,000 ÷ 51 = 196 tokens └─ Virtual Agent: ~1,000 sessions/day └─ 1,000 × 20 × 2 = 40,000 tokens Total Monthly: ~93,527 tokens Included: 250 × 1,000 = 250,000 tokens Overage: None Cost: $0 overage Note: This enterprise center is well within allocation even with heavy AI usage across all capabilities ``` --- ## Token Consumption Monitoring ### Token Usage Dashboard ``` Organization Token Summary Period: March 2026 Total Allocation: 50,000 tokens/month Total Consumed: 19,810 tokens Remaining: 30,190 tokens (60% available) Usage Rate: 40% Consumption by Feature: ├─ Agent Copilot: 11,000 tokens (55%) ├─ Virtual Agent: 8,000 tokens (40%) ├─ AI Scoring: 500 tokens (3%) ├─ Voice Bots: 294 tokens (1.5%) ├─ Messaging: 13 tokens (0.1%) ├─ AI Translate: 3 tokens (0.1%) └─ Other: 0 tokens Consumption Trends: ├─ Week 1: 4,500 tokens (15% of monthly) ├─ Week 2: 4,200 tokens (14% of monthly) ├─ Week 3: 5,300 tokens (18% of monthly) ├─ Week 4: 5,810 tokens (19% of monthly) └─ Trend: Stable, slight increase toward month-end Projected Usage (if trend continues): └─ End of Month: 19,810 tokens (40% of allocation) Health Status: ✓ GREEN (well within limits) Recommendation: Can safely increase AI usage ``` ### How to Access Token Monitoring 1. Log into Genesys Cloud Admin 2. Navigate to Reports → Usage 3. Select "AI Token Consumption" 4. Choose time period to review 5. View consumption by feature 6. Export data if needed ### Key Metrics to Track | Metric | Purpose | Good Range | |---|---|---| | Overall Utilization | Total token usage | 40-80% of allocation | | Per-Feature Usage | Identify heavy consumers | Proportional to use | | Trend Direction | Usage pattern | Stable or controlled | | Overage Risk | Approaching limits | Monitor if >80% | | Unused Allocation | Wasted capacity | <20% unused | --- ## Cost Management Strategies ### Optimization Techniques - **Improve Bot Deflection** - Reduce voice bot minutes through better automation - **End Sessions Properly** - Digital bot sessions end cleanly to reduce count - **Optimize Routing** - Use Predictive Routing (included) instead of manual methods - **Batch Processing** - Group analytics and translations for efficiency - **Selective Features** - Enable only needed AI features per queue - **Clean Knowledge Base** - Reduce unnecessary knowledge article access ### Budgeting Approach 1. **Baseline Calculation**: - 50 tokens per Agent Copilot user/month - 2 tokens per Virtual Agent session - 1 token per 20 evaluations - 1 token per 51 digital bot sessions 2. **Add Buffer**: - For unpredictability: +20% overage buffer - For growth: +10-15% growth buffer 3. **Monitor Monthly**: - Review actual vs. projected - Identify variance - Adjust next month's budget 4. **Planning**: - Budget for peak season - Factor in new initiatives - Plan for scaling ### Real-World Budget Example ``` Mid-Market Center - Annual Budget Planning: Current State (March 2026): ├─ Monthly consumption: 19,810 tokens ├─ Monthly allocation: 50,000 tokens ├─ Utilization: 40% └─ Cost: $0 overage Planned Growth (Next 12 months): ├─ +50 agents (10% growth) ├─ New Virtual Agent deployment ├─ Expanded Messaging channels └─ Expected consumption increase: +8,000 tokens/month Future State Projection: ├─ Monthly consumption: ~27,800 tokens ├─ Monthly allocation: 62,500 tokens (250 × 250 users) ├─ Utilization: 44% └─ Cost: $0 overage Annual Cost Impact: ├─ Base licensing: No increase (same CX tier) ├─ AI Tokens: No additional cost (within allocation) └─ Total Additional Cost: $0 (internal reallocation only) Conclusion: Growth sustainable within existing token allocation ``` --- ## Token Pricing by Region ### Pricing Variations - **North America**: ~$1.00 per token - **Europe**: Varies by currency/region - **APAC**: Pricing adjusted for region - **Other Regions**: Contact Genesys for pricing ### Ordering Additional Tokens - **Option 1**: AppFoundry - Self-service purchasing - **Option 2**: Genesys Representative - Volume discounts available - **Payment**: Monthly billing or prepaid options - **Minimum**: No minimum purchase requirement - **Flexibility**: Increase or decrease allocation monthly --- ## Real-World Usage Scenarios ### Scenario 1: Seasonal Spike Management ``` Contact Center Business: Holiday Retailer Challenge: Heavy AI usage Dec-Jan, minimal Jun-Aug Strategy: ├─ Base allocation: 30,000 tokens/month ├─ November: Monitor usage, prepare for peak ├─ Dec-Jan: Purchase additional 20,000 tokens │ └─ Ensures sufficient capacity │ └─ Cost: ~$20,000 for 2 months ├─ Feb: Return to base allocation └─ Jun-Aug: Only use allocation (excess unused) Cost Management: ├─ Off-peak: $0 additional cost ├─ Peak months: $20,000 additional ├─ Annual cost: $20,000 (2 months only) └─ Flexibility: Scale as needed ``` ### Scenario 2: Phased AI Rollout ``` Contact Center: Implementing AI progressively Phase 1 (Month 1): Agent Copilot only ├─ Consumption: 4,000 tokens (10 users) ├─ Cost: $0 (within allocation) └─ ROI: Proven before major investment Phase 2 (Month 3): Add Virtual Agent ├─ New consumption: +3,000 tokens ├─ Total: 7,000 tokens ├─ Cost: $0 (still within allocation) └─ Result: Expand automation safely Phase 3 (Month 6): Add Analytics + Bots ├─ New consumption: +5,000 tokens ├─ Total: 12,000 tokens ├─ Cost: $0 (allocated) └─ Result: Full AI platform Phase 4 (Month 12): Scale based on ROI ├─ Purchase additional tokens if needed ├─ Cost: Only what you use └─ Confidence: Proven ROI before major spend ``` --- ## Interview Cheat Sheet | Question | Answer | |---|---| | What is token model? | Consumption-based pricing for AI features | | What's the default allocation? | 250 tokens per named user, 350 per concurrent user | | What's token cost? | ~$1.00 per additional token beyond allocation | | How is consumption tracked? | Real-time metering per feature with monthly reporting | | Which features consume tokens? | Agent Copilot, Virtual Agent, Bots, Translate, Analytics, etc. | | Agent Copilot consumption? | 40-60 tokens per user per month | | Virtual Agent cost? | 2 tokens per session | | Voice Bot cost? | 1 token per 17 minutes | | Digital Bot cost? | 1 token per 51 sessions | | AI Scoring cost? | 1 token per 20 evaluations | | Does Predictive Routing cost? | No, included with platform | | How do I monitor usage? | Admin → Reports → Usage → AI Token Consumption | | Can I purchase additional? | Yes, through AppFoundry or Genesys rep | | How often do tokens renew? | Monthly, with no carryover | | Is there overage protection? | Fair use policy covers most organizations | | What about fair use? | 95% of customers covered by fair use policy | | How to optimize costs? | Improve bot deflection, optimize routing, batch processing | | Can I predict usage? | Yes, based on user count and AI feature deployment | --- ## Key Takeaways - **Flexible Pricing** - Pay only for what you use - **Default Allocation** - All CX tiers include base tokens - **Transparent Metering** - Each feature has clear consumption rates - **Cost Control** - Budget and monitor usage in real-time - **Scalability** - Add tokens as needs grow - **Fair Use** - 95% of customers within standard allocation - **No Overage Risk** - Purchase additional tokens as needed - **Monthly Billing** - Consumption tracked and billed monthly - **Optimization Available** - Multiple strategies to reduce costs - **Regional Flexibility** - Pricing adjusted by region --- ## Cost Calculation Tool ### Quick Estimation Template ``` Contact Center Token Cost Calculator: 1. Agent Copilot Users: ____ × 50 tokens = _____ tokens 2. Virtual Agent Sessions/Day: ____ Daily calculation: ____ sessions × 2 tokens = _____ tokens/day Monthly: _____ tokens/day × 20 days = _____ tokens 3. AI Scoring Interactions/Day: ____ Monthly: (____ × 20 ÷ 20) = _____ tokens 4. Voice Bot Minutes/Day: ____ Monthly: (____ × 20 ÷ 17) = _____ tokens 5. Digital Bot Sessions/Month: ____ Monthly: (____ ÷ 51) = _____ tokens 6. Messaging/Month: ____ messages Monthly: (____ ÷ 400) = _____ tokens 7. Other Features (translate, analytics, etc): _____ tokens TOTAL MONTHLY CONSUMPTION: _____ tokens Your Allocation: 250 × (number of users) = _____ tokens Overage: (Total - Allocation) or $0 if under Annual Cost: ├─ Base: Included in licensing ├─ Overage: (Monthly overage × 12) × $1.00 └─ Total: _____ annually ``` --- ## Additional Resources ### Official Documentation - Token-Based Pricing: help.genesys.cloud/articles/genesys-cloud-tokens-based-pricing-model/ - AI Token Billing: help.genesys.cloud/articles/ai-token-billing/ - AppFoundry Tokens: appfoundry.genesys.com (search "AI Experience Tokens") - Pricing Overview: genesys.com/pricing ### Support & Ordering - Genesys Sales: sales@genesys.com - Genesys Support: https://support.genesys.com - AppFoundry: https://appfoundry.genesys.com - Community Forums: https://community.genesys.com --- ## Document Version Info **Last Updated:** March 2026 **Source:** Genesys PureCloud Official Documentation **Validated:** Current with January-March 2026 releases **Version:** 1.0