Salesforce launches automated AI Agent testing center

Salesforce graphic of an AI agent in front of a software screen and graphic mountains and trees in the background

Key Takeaways

Salesforce has launched agentic lifecycle management tools to automate Agentforce testing and manage AI agents without disrupting live production.

The new testing center enables enterprises to test, deploy and monitor AI agents at scale.

The introduction of AI-generated synthetic test interactions and enhanced Data Cloud Sandboxes will allow companies to accurately test AI agents with minimal risk while optimizing costs.

Global CRM leader Salesforce has unveiled agentic lifecycle management tools to allow users to automate Agentforce testing, prototype agents in secure Sandbox environments and transparently manage usage at scale.

According to the spokespersons, the demand for the toolchain grew from the need to test and configure AI agents without disrupting live production environments. The testing center will therefore enable teams to test, deploy and monitor AI agents with Agentforce at scale, enabling every enterprise to become “agent-first.”

Some of the capabilities that users should look forward to include AI-generated test interactions which will allow customers to test Agentforce using synthetically generated data, ensuring accurate responses and actions with monitoring of usage and feedback.

Additionally, New Data Cloud Sandboxes and Digital Wallet enhancements will enable teams to test the entire AI and data lifecycle in secure test environments while transparently managing and optimizing costs.

“Agentforce is helping businesses create a limitless workforce. To deliver this value fast, CIOs need new tools for testing and monitoring agentic systems,” said Adam Evans, EVP and GM for Salesforce AI Platform. 

“[…] This new category of Agentic Lifecycle Management requires unique tools, and Salesforce is meeting the moment again with Agentforce Testing Center, which will help companies roll out trusted AI agents with no-code tools for testing, deploying and monitoring in a secure, repeatable way.”

Providing an analytical insight, Keith Kirkpatrick, research director at The Futurum Group, shared:

“To help engender trust in AI, every business investing in an agent strategy will need to think through how they pressure test an autonomous AI agent that can reason, retrieve data and use tools. Salesforce is instilling confidence by using AI to test hundreds of variations of an agent’s interaction in parallel. 

“With these enhancements and support for Agentforce and Data Cloud in familiar testing environments, this Testing Center will help Salesforce customers manage the unique needs of testing AI agents.”