The issue was caused by a disruption in a downstream provider, but this only affected our uptime check. No customer was impacted.

The Modern Treasury website has recovered.

This incident has been resolved.

We've identified the problem. Our canary tests are affected, but no customers are impacted.

The Modern Treasury API latency has recovered and all operations are normal.

This incident was caused by resource contention in our shared database cluster that started at around 0100 PT. Our on-calls were paged within minutes, and began investigating the issue.

Our API latencies began sporadically exceeding our internal SLOs at 0200 PT, and then were consistently above our SLOs at 0700 PT. Ultimately, the incident response team resolved the performance degradation through a combination of workload management and query optimizations. The API performance returned to SLO levels at around 0950 PT.

The Modern Treasury API latency has recovered. We are currently monitoring.

This incident has been resolved. We have determined that it was a false positive, and production API calls were not affected.

This incident has been resolved.

We just resolved the issue!

From 12:25 am PT to 12:38 am PT, Plaid was down for maintenance, causing API requests to Plaid to fail.