Phase 08Trust, cost and visibility427 of 434
Observability for something you cannot reproduce
Trace every call with its inputs, its tokens, its latency and its outcome, because you will not reproduce the bad answer twice.
Build40 minAI implements
Build this in your own editor
This one runs on your machine rather than in the browser workbench. Work to the outcomes below, then come back and mark it complete.
You should now be able to
- Trace a multi-step AI request end to end
- Record tokens, cost and latency per span
- Get from a user complaint to the exact request in under a minute
Loading…