Gong launches Agent Builder to automate revenue-signal driven sales actions

Gong announced the Agent Builder feature of its Revenue AI platform, enabling sales and RevOps teams to automatically construct agents that convert buyer signals into concrete actions link. The tool connects signals to the appropriate account, opportunity, and sales stage, then routes output to systems like Slack, email, or CRM fields.

A use case described involves a buyer expressing interest in a product they don’t currently use; the custom agent identifies the interest, drafts outreach, sends a Slack alert to the rep, and creates a review task, preventing the signal from being missed. Riskonnect’s Director of Sales Enablement, Brian Seitz, said the agent helped surface additional needs and expand conversations beyond the initial solution.

Agents draw on the Gong Revenue Graph for contextual data and run under the Revenue Harness, which enforces permissions, guardrails, and cost-aware execution. Incremental updates process only changed data, reducing unnecessary computation.

Output actions are executed through Gong Applications or extended via the Model Context Protocol (MCP), allowing updates to CRM fields, contact enrichment, flow initiation, and more, keeping the insight within the teams’ existing workflow.

The announcement positions Agent Builder as a bridge from experimental AI agents to production-grade revenue execution, promising more consistent sales processes and clearer ROI on AI investments.