2026-10-12¶
OllyGarden¶
Congrats to Juraci Paixão Kröhling and team: OllyGarden turbo charges mission to fix bad telemetry with $4m raise and launch of “Minimum Viable Instrumentation”.
AI agent failures¶
In How to detect silent AI agent failures in production Armend Avdijaj provides a framework and concrete tips on a typical prod issue.
OTel logs bridge¶
I Built an OpenTelemetry Logs Bridge. Encoding Was the Easy Part by Admilson Cossa is a interesting hands-on deep dive.
Ops in the age of AI¶
Together with Sujana Sooreddy of Netflix and Noam Levi of Groundcover, yours truly was on a virtual InfoQ panel, the recording and transcript is now available: Beyond Observability: Evolving Production Operations in the Age of AI.
Tooling¶
Tools I stumbled upon and found useful:
- bronto-community/compy is a local OTel collector manager for macOS including nice UI and powerful CLI.
- llm-measurement/otelcol-genai-sketches allows to track LLM token usage by user, session, and model in Prometheus or via OTLP ingest.
That was it for this edition of the o11y newsletter. Feel free to share news items with me via Bluesky DM. Stay safe and hope to see you around next week!