Agentic virtual agents overview

AI-generated summary

Genesys Cloud AI Guides enables business users to create AI-powered Virtual Agents using natural language without coding. The system features natural language instructions for building conversational flows, dynamic adaptation to customer context, and blended structured flow logic with AI. Users can support task completion, data gathering, and human agent handoffs while enabling rapid iteration without developer involvement. The upcoming Large Action Model (LAM) feature in AI Studio allows for fully agentic virtual agents that can reason across actions and knowledge, combining AI flexibility with structured logic through Architect deployment. Key benefits include action-oriented AI agents, reduced hallucination risk compared to traditional LLMs, improved conversational experiences, higher self-service effectiveness, and lower authoring effort. The system operates on a token-based pricing model for interaction sessions, requires the Genesys Cloud AI Experience license, and integrates with Architect Virtual Agent flows for enterprise-scale deployment. Configuration is managed through AI Studio's GUI for setting capabilities, tools, knowledge, and compliance guardrails.

Feature coming soon: Agentic virtual agents

Large Action Model (LAM)–powered virtual agents in AI Studio enable business teams to design fully agentic, conversational, and action-oriented virtual agents that can reason across actions and knowledge. These agents combine the flexibility and natural interaction of modern AI with structured logic and governance, allowing safe, scalable deployment through Architect. By reducing reliance on rigid bot flows and minimizing hallucination risks associated with traditional LLMs, LAM-based agentic virtual agents deliver more reliable, on-brand, and effective self-service experiences.

Key benefits

Action-oriented AI agents: LAMs are designed to act as instructed, enabling virtual agents to reason, act, and complete tasks reliably.
Reduced hallucination risk: Compared with general-purpose LLMs, LAMs significantly lower the likelihood of hallucinations, increasing trust and business confidence.
Improved conversational experience: More natural, engaging conversations reduce friction and create smoother user interactions.
Higher self-service effectiveness: Increases key metrics such as containment, first contact resolution, and task completion rates.
Lower authoring effort: Covers more use cases with fewer manual instructions compared with guides.
Safe and scalable deployment: Combines agentic AI with structured logic in Architect, ensuring governance, scalability, and enterprise readiness. AI Studio provides a GUI to configure capabilities, tools, knowledge, and guardrails, ensuring consistent and compliant agent behavior.

Get started

Pricing policy

Genesys Cloud AI Experience tokens help you monitor and manage feature consumption and offer flexibility for changing business needs. Agentic virtual agents consume tokens for each interaction session. However, no tokens are charged for the creation of agentic virtual agents. For more information, see Genesys Cloud tokens-based pricing model.