Status

One service is degraded.

Partial service degradation
status.memoryfirst.ai · 99.97% rolling 90-day uptime
Updated 11 s ago
Infrastructure graph · liveRefreshes every 60s · last 11s ago
msa-acme-2025consult-2026-03-consult-2026-02-docspricing-2026deploy
Component status

Every service that ships your answers.

90-day uptime windows below. We publish per-service rolling SLA, not just the overall figure.

API · /v1/query
247 ms p50
99.99% · 90d
operational
MCP server
42 ms p50
99.97% · 90d
operational
Embeddings
38 ms p50
99.99% · 90d
operational
Memory store · primary
8 ms p50
100% · 90d
operational
Object store · audio
21 ms p50
99.95% · 90d
operational
Web crawler workers
queue backed up · 4 min
99.21% · 90d
warn
Authentication (SSO)
18 ms p50
99.99% · 90d
operational
Billing · Stripe (EU)
upstream healthy
depends · 90d
operational
Recent activity

The last 30 days, unfiltered.

May 19, 2026 · 14:33 CET
duration · 38 min
minorresolved

Web crawler queue backed up after CDN edge config change

Cloudflare edge config update caused crawler requests to retry against stale origins. Caught by alerting at 14:35, rolled back at 15:11. No ingested data lost; affected crawls re-queued.

May 04, 2026 · 02:14 CET
duration · 12 min
minorresolved

Memory store replica lag during scheduled vacuum

Scheduled full vacuum on the EU-FAL1 cluster caused 12 minutes of >2s p99 on read replicas. Queries automatically failed over to primary. No customer-visible errors.

Apr 22, 2026 · 09:00 CET
duration · 4 min
maintenancecompleted

Planned maintenance · Embeddings model upgrade to 1.2.0

Rolled embeddings model from 1.1.4 → 1.2.0. Re-embedded all docs in the background (no downtime). p50 latency dropped from 47 ms → 38 ms post-upgrade.

Subscribe to status updates.

Email, RSS, or a webhook into your own ops channel. We post the first acknowledgement within five minutes of every incident.