build1 publisher
Foundry's automated grading needs the trace to carry the content it grades
Microsoft Foundry writes agent runs as OpenTelemetry spans with gen_ai attributes and exports them to Azure Monitor, where evaluators score sampled production traffic off the same records. Those attributes include tool arguments.
Publishers:dev.to
Reality
- Evidence40
- Adoption
- Insufficient
- Hype gap+10
- Incentives55
- Confidence45