Summary

On April 21 and 22, a partial outage at one of our upstream providers prevented roughly 30% of customers from logging in or accessing the platform. The issue has been fully resolved.

Impact

Affected services: Platform and API
Unaffected services: Website and customer support
Scope: ~30% of traffic
Duration: Three incidents of ~45 minutes each — two on April 21 and one on April 22

Cause

An outage at one of our infrastructure providers disrupted a subset of our traffic, causing login and platform access failures for customers routed through the affected segment.
Resolution & preventative measures
The underlying issue was resolved on April 22. To reduce the likelihood and blast radius of similar incidents, we have:

Scaled our servers to 6 replicas for greater redundancy
Added improved detection to catch provider-level issues sooner
Introduced fallback processes that isolate failures to a specific API, rather than impacting all customers on an affected server

We apologise for the disruption and appreciate your patience.