Operation is back to normal.

We have identified the failing component and are deploying mitigations.

We are investigating a problem regarding run initialization.

The issue is resolved, and mitigations are in place to avoid this issue in the future.

We've identified a problem with an internal service and are deploying mitigations.

We have deployed the fix, and William should now be working as intended again.

We have identified the issue and are working on deploying a fix.

We are currently experiencing an issue where William is not responding to user messages, resulting in an error message. We are investigating the issue.

The elevated error rates from our LLM provider have been resolved, and runs are operating normally again.

We’re seeing elevated error rates from one of our LLM providers, causing some runs to fail with MODEL_SERVICE_ERROR.

We identified and resolved an issue affecting one of our infrastructure pieces that lead to some runs not starting properly.

We are investigating some tests not starting as expected, resulting in errored runs.

The affected service has been restored and is operating normally again.

Our service provider is currently experiencing a partial outage.

Test and Testplan runs are fully operational.

We have identified the issue: one of our VMs ran out of memory and could not automatically recover. We will restore it. This affects start of new runs, including testplans, and might degrade in-progress runs.

Errors about testplans not starting have surfaced.