Evaluate Entire Conversations for Voice and Chat Agents with Testing Center in Agentforce Studio (Beta)

Simulate and score full multi-turn conversations at scale to validate that your voice and chat agents resolve customer issues end-to-end. Generate up to 20 cases or upload up to 1,000 test cases, run them against realistic customer personas, and get scored results with transcripts and audio recordings.

Where: This feature is available in Lightning Experience in Enterprise and Unlimited editions with Agentforce.

Note Conversation-level testing is a pilot or beta service that is subject to the Beta Services Terms at Agreements - Salesforce.com or a written Unified Pilot Agreement if executed by Customer, and applicable terms in the Product Terms Directory. Use of this pilot or beta service is at the Customer's sole discretion.

When: This change is available starting the week of August 17, 2026.

Why: An agent can score well on every individual response and still fail the customer. Conversation-level testing closes that gap by automatically simulating and scoring full conversations at scale, so you can catch end-to-end failures before deployment without manually reviewing every interaction.

How: In Testing Center in Agentforce Studio, create a test suite and select Conversation-Level Testing. On the Data step, select Text only or Text and voice for the evaluation mode, choose a customer persona, then upload a CSV file or generate test cases.

Task resolution is automatically included as an evaluation and scores conversations from 1 to 5 on goal completion and intent alignment. Results include a transcript, a task resolution score with reasoning, and an audio recording of the simulated conversation for voice agents.

Conversation-level testing supports a maximum of 3 concurrent test suites and 20 turns per conversation. You can generate up to 20 test cases from an LLM or upload up to 1,000 test cases from a CSV file.