Post Content
Integrating evaluations is critical to iterate on agent configuration including for model upgrades, system prompts and model parameters in production. In this demo we’ll showcase how you can leverage evaluations initiated in CI/CD workflows to make changes to your Agent configuration with confidence. Once in production you can use continuous evaluation in the Foundry Observability dashboard to monitor the quality and safety of your Agent and leverage tracing for in-depth debugging.
To learn more, please check out these resources:
* https://aka.ms/build25/plan/CreateAgenticAISolutions
𝗦𝗽𝗲𝗮𝗸𝗲𝗿𝘀:
* Han Che
* Ami Patel
𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻:
This is one of many sessions from the Microsoft Build 2025 event. View even more sessions on-demand and learn about Microsoft Build at https://build.microsoft.com
DEM528 | English (US) | AI, Copilot & Agents
#MSBuild
Chapters:
0:00 – Setting up Tracing and Monitoring with Application Insights
00:01:51 – Configuring Azure Monitor and Starting Instrumentation
00:02:09 – Continuous Evaluation in Production for Agent Efficiency
00:03:55 – Default Settings and Customization Options for Reasoning Scores
00:04:21 – Monitoring and Analysis in Azure AI Foundry
00:06:02 – Identifying and Diagnosing Performance Issues
00:07:19 – Introduction to AI Agent Evaluation Extension
00:08:42 – Baseline and Comparative Agent Evaluation
00:11:20 – Introduction to Operational Matrix Read More Microsoft Developer
