Announcedmicrosoft-foundryEvaluation & observabilityAgent governanceDay 1

Continuous improvement comes to Microsoft Foundry

Foundry gains a continuous improvement loop: agent traces and expert feedback become evaluations, and an agent optimizer proposes tested changes to instructions, skills and model choice. Insights and voice agents are generally available, while skill optimisation, cost optimisation and Programmed Training are in preview.

Recap by Jussi Roine · Updated 17 Nov 2026
status
Announced
why it matters

Agents that change themselves over time are a governance question as much as a quality one. Microsoft’s answer is human approval: nothing changes in production until the team applies it, and skills are versioned for rollback. Most of the loop is still in preview, and the customer cost savings quoted are vendor-supplied.

Evaluation & observabilityAgent governance

Official sources

Community deep dives

0 links · newest first
Written or recorded something about this?It goes to the moderators with this recap attached. No account needed, and you can get the credit.
Add a link to this recap