Changes / Count / Degraded AI and API operations
Count: Major incident: Degraded AI and API operations
Summary
Count reported major degradation of AI, private/public APIs and query execution across EU and US components, starting 2026-06-19 15:32 UTC and resolved at 2026-06-19 15:45 UTC (12 min). Count suspected private GCP endpoint access issues, identified a potentially responsible infrastructure update and rolled back to a last-known-good state.
Machine-written from the diff below (Sol; type: incident degradation). Check the diff before relying on it.
What changed
Before: not present
After: Degraded AI and API operations
Before: not present
After: major
Before: not present
After: resolved
Before: not present
After: 2026-06-19 15:32 UTC
Before: not present
After: 2026-06-19 15:45 UTC
Before: not present
After: EU - Private API (EU), EU - Public API (EU), EU - Query execution (EU), Private API (EU), Private API (US), Public API (EU), Public API (US), Query execution (EU), Query execution (US), US - Private API (US), US - Public API (US), US - Query execution (US)
Before: not present
After: Several of Count's backend services related to AI and private/public API operations are currently in a degraded state. We are currently investigating the cause, we believe it is related to issues accessing private GCP endpoints.
Before: not present
After: We have identified an infrastructure update which may have caused the issue - we have rolled back to a last-known-good state and are continuing to investigate the cause.
Before: not present
After: This incident has been resolved.
First version of this incident archived by Rewindex. View the live incident.
History of this incident
| When (UTC) | Change | Kind |
|---|---|---|
| Major incident: Degraded AI and API operations (this change) Count reported major degradation of AI, private/public APIs and query execution across EU and US components, starting 2026-06-19 15:32 UTC and resolved at 2026-06-19 15:45 UTC (12 min). Count suspected private GCP endpoint access issues, identified a potentially responsible infrastructure update and rolled back to a last-known-good state. | Major incident |