Salesforce adds Agent Evals and cost controls to Agent Fabric for enterprise AI
Salesforce announced new features for Agent Fabric on October 27, 2026, targeting enterprise AI governance. The platform, which handles over 1.2 billion daily requests, now includes Agent Evals to automatically test agents and models in production, triggering interventions like the Agent Kill Switch when quality drops. API Evals detects breaking API changes before they reach production, while…
Key points
- Agent Fabric now auto-tests agents and models in production, triggering interventions if quality falls below thresholds
- Cost Controls enforce budget limits, cutting token usage by up to 75% via semantic caching and response trimming
- API Evals detects API changes before they break agent workflows, with general availability in January 2027
The updates follow survey data from LangChain’s 2025 report, which found only 37.3% of organizations evaluate agents on live traffic. Salesforce cites customer feedback, including from ASICS Digital, Capita, and GE Appliances, highlighting demand for automated quality checks and budget enforcement. Most features are now generally available, with Agent Evals and Custom Agent Scanner entering beta in November 2026 and full release in January 2027.
Salesforce Adds Agent Evals and Cost Controls to Agent Fabric
Unite.AI · 27 October 2026
Loading the full article…
This text was published by Unite.AI and written by Aiden Cross, AI Product Strategy & Execution, AI Research Agent. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Enterprise & Industry
All →- Meta AI director Rohit Patel launches Ndoc to replace Microsoft Office and Google Workspace · 1 src
- Simplismart raises $14M Series B to optimize AI model deployment · 1 src
- VeriPark earns Microsoft’s 2026–2027 AI Business Solutions Inner Circle award · 1 src
- ServiceNow and Google Cloud link Gemini Enterprise to industry workflows via Action Fabric · 1 src
- Microsoft Claims Its New Medical AI Diagnostic Orchestrator Is Better at Diagnosing Than Real Doctors · 18 src
Comments
via GitHub Discussions