Genesys cloud agentic virtual agent unveiled

Genesys has introduced the Genesys Cloud Agentic Virtual Agent, describing it as the first enterprise customer experience solution built on large action models to enable autonomous, end-to-end resolution of customer requests.

The announcement marks a shift from traditional self-service models toward systems designed to interpret goals, determine next steps and execute actions across front and back-office environments.

From conversational AI to action-driven systems

Many organisations have adopted virtual agents based on large language models to improve conversational quality. However, these systems are typically limited to generating responses and often struggle with multi-step workflows spanning multiple systems.

According to a Gartner report cited in the announcement, the average self-service success rate stands at 22 percent, while 46 percent of surveyed leaders rank improving self-service as a top-three priority for 2026.

The Genesys Cloud Agentic Virtual Agent aims to address these gaps by combining large action models with orchestration and governance capabilities. The system is designed to move beyond answering queries and instead complete tasks across enterprise systems.

How the model operates

The virtual agent serves as an orchestration layer capable of progressing workflows across CRM, billing and service operations platforms. It is built to:

• Understand customer goals and context

• Plan multi-step actions

• Execute tasks across systems

• Adapt as conditions change

• Maintain explainability and auditability

The platform integrates the Scaled Cognition APT-1 large action model, designed for deterministic and action-grounded execution. According to the company, this reduces hallucinations and improves reliability compared to text-focused models.

Embedded guardrails and policy controls aim to ensure that actions remain predictable and aligned with enterprise requirements.

Enterprise exploration underway

Several organisations are already exploring the capability, including M&T Bank, Banco Pichincha, a Fortune 500 healthcare company and a Fortune 50 North American retailer.

Olivier Jouve, Chief Product Officer at Genesys, stated that autonomy in customer experience must be built on trust and control, adding that the system is designed to reason, plan and act across systems while maintaining governance.

Dan Roth, Co-founder and CEO of Scaled Cognition, said that traditional large language models are designed to generate text rather than execute tasks, creating reliability gaps in enterprise automation.

Hayley Sutherland, Research Manager for Conversational AI at IDC, commented that resolving complex customer requests requires AI capable of planning and executing multi-step actions while remaining predictable and auditable.

Trond Prestø, Head of Customer Care at DNB, said agentic AI represents the next stage in scaling customer resolution while maintaining governance.

Governance and compliance focus

The Genesys Cloud Agentic Virtual Agent extends the platform’s governance model with action-level explainability and audit trails. Through Genesys Cloud AI Studio, organisations can configure guardrails, permissions and behavioural controls aligned with internal policies.

The company also plans to support open standards such as Agent-to-Agent and Model Context Protocol, enabling coordination with other AI agents and enterprise systems.

Availability timeline

The Genesys Cloud Agentic Virtual Agent, powered by large action models, is expected to be generally available globally in the first quarter of the company’s fiscal year 2027, covering 1 February 2026 to 30 April 2026.

The launch reflects an industry shift toward automation models that not only respond to queries but complete tasks across enterprise environments, with governance embedded at each stage.