Ops & SLO

What is this?

SLO (24h): average and p95 latency, low‑evidence rate computed from recent eval logs.
Status: backend configuration flags (OTEL, S3, API key, org requirement).
Calibration: recalculates relevance threshold used by analytics/QA dashboards.
Self‑Heal: runs index/schema maintenance tasks; safe for production.
Compact/Dedup: removes duplicate chunks and reclaims storage.

SLO (24h)

Total: -

Avg latency: - ms

p95 latency: - ms

Low-evidence rate: -

Status

OTEL: off

S3: off

API key: off

Require Org: no

Calibration

Threshold: -

Last run: -

Actions