Author: Microsoft Developer - Bewertung: 2x - Views:12
Nondeterministic, multi-agent systems break traditional monitoring. As agents reach production, observability must be built in—not added after failures. This session covers modern agent observability: cross-framework tracing and evals, rigorous inner-loop practices, evolving context-specific evals, and always-on signals that connect behavior to business outcomes to measure value, cost, and ROI.
Seating for this session is first-come, first-served. Add it to your schedule to plan your day and arrive early to secure a spot.
To learn more, please check out these resources:
* https://aka.ms/build26/BRK252
* https://aka.ms/build/foundrydiscord
𝗦𝗽𝗲𝗮𝗸𝗲𝗿𝘀:
* Sebastian Kohlmeier
* Filisha Shah
* Vivek Bhadauria
𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻:
This is one of many sessions from the Microsoft Build 2026 event. View even more sessions on-demand and learn about Microsoft Build at https://build.microsoft.com
BRK252 | English (US) | Responsible AI
Breakout | (300) Advanced
#MSBuild
Chapters:
0:00 - Integration with Foundry’s control plane and building reliable production agents
00:05:03 - Demonstration of agent evaluation process and introduction to rubric evaluator
00:10:31 - Foundry Integrated Observability with Azure Monitor
00:11:53 - Transition to Agent DevOps Loop and Code-First Observability
00:13:40 - Using Rubric Evaluator for Multi-Dimensional Analysis
00:17:43 - Overview of VS Code, GitHub Copilot, and CLI capabilities
00:19:17 - Walkthrough of evaluator catalog and multi-turn evaluators
00:26:19 - Continuous Agent Optimization and Iterative Improvement
00:34:21 - Upcoming Sessions, Resources, and Closing Remarks
SOCIAL SHARE CARD GENERATOR