Reference · Glossary
AI monitoring
Last updated
Any production LLM or ML endpoint. Pair infra signals (latency, 5xx) with quality signals (eval samples, user thumbs-down, citation fail rates).
#When to use
Any production LLM or ML endpoint. Pair infra signals (latency, 5xx) with quality signals (eval samples, user thumbs-down, citation fail rates).
#When not to
Shipping without a baseline eval. Dashboards without action thresholds become wallpaper.
#Quality checklist
- Golden-task sample scored on a schedule
- Alerts on cost spikes and error budgets
- Trace IDs linking prompt → retrieval → completion
- Owners for quality vs infra pages
#Learn next
- Lesson: `ai-monitoring`
- Related: observability, eval set