Issues with clearing, editing learner progress and forcing retakes in Learn365

Update timeline

1. identified Jun 05, 2026, 02:19 PM UTC

   We are currently experiencing an issue affecting learner progress management actions in Learn365. Some customers may be unable to clear learner progress, edit learner progress, or force a retake. These actions may keep loading without completing or return an error. Our engineering team has identified the issue as related to a recent migration and is actively working on a fix. We expect the fix to be available soon. We will provide further updates as soon as more information is available.

2. resolved Jun 05, 2026, 03:11 PM UTC

   The fix has been deployed, and the affected functionality should now be working as expected. If you experience any unexpected behavior, please reach out to our Support team by submitting a ticket: https://helpcenter.zensai.com/hc/en-us/articles/14197648878237-How-to-contact-Zensai-Product-Support

3. postmortem Jun 09, 2026, 04:13 PM UTC

   ### **Root Cause and Analysis** During the **Learn Service deployment**, a database migration script removed a column that was still actively referenced by the monolith service. The Learn Service deployment should have been performed only after a prior monolith deployment. That prior deployment would have removed the remaining dependency on the column. However, this deployment order dependency was not formally documented or communicated to the Release Manager before the release. In addition, an established but undocumented engineering best practice was not followed: **database column removal should be done in a separate, later release after the column has already been removed from the application entity**. Removing the column within the same release increased the risk of compatibility issues between services. Several contributing factors also increased the likelihood of human error, including: * insufficient cross-team communication around deployment order dependencies; * a high number of concurrent work items; * delayed Staging validation due to a significant regression backlog; * reduced QA capacity due to team members being on leave. ### **Resolution** Once the root cause was identified, the rollback part of the database migration script was extracted and applied as a **hotfix** to the Learn Service. This restored the removed database column and re-established compatibility with the monolith service. After the hotfix was applied, the affected functionality was restored. ### **Remediation Items** To reduce the risk of similar incidents in the future, we are implementing the following improvements: **Improved deployment dependency communication** Deployment order dependencies must be explicitly documented and communicated to the Release Manager before any deployment request. **Formalized database migration guidelines** The rule that database column removal must be separated from application-level entity changes into a later release will be added to official engineering guidelines and release checklists. **AI-assisted deployment review** An AI agent has been developed and deployed to help detect and flag potentially risky migration patterns, including column removals that may still have active cross-service dependencies. **Stronger release readiness gates** Additional validation checkpoints will be introduced to ensure Staging sign-off and critical release checks are not bypassed or delayed due to workload or resource constraints. We apologize for the disruption this caused and appreciate your patience while our teams investigated and resolved the issue.