Status · incident history
Last 90 days of incidents.
We post every incident publicly, even the embarrassing ones. Root cause, impact, duration, remediation. Want to be notified the moment one starts? Subscribe via Slack, webhook, or RSS.
2026-04-05
Major
Duration: 47 min
Brief generation paused for ~50 minutes
Upstream LLM provider returned 429s sustained for 47 minutes; we didn't have a fallback wired correctly. Switched primary provider; backfilled the 2,100 stalled briefs by 18:00 UTC. Read the full post-mortem →
No other incidents in the last 90 days
Live uptime + current incident state is on the status page.
Our incident discipline.
Detection target: sub-5-minute detection for outages affecting brief generation, signal detection, or any push-to-CRM path.
Public-post target: within 15 minutes of confirmation. Even partial information beats silence.
Post-mortem target: published within 5 business days, including the dumb-mistake version of what happened, not just the engineering-flavored one.