Purpose
Incidents are contained quickly, customers hear about them from us, and each one leaves the service better than before.
Detection
- Automated health checks of the app, the API, the database and the website, with alerts to the operators.
- Daily quality checks of the model routes kanman pays for; a route whose answers get worse stops being used for kanman’s own steps.
- Reports from customers and researchers to [email protected].
Containment
Containment comes first. The tools for it are:
- Stop all runs for a workspace, which stops every run within a minute, in the cloud and on runners.
- Stopping a model route or a run machine version for every team at once.
- Revoking tokens and rotating credentials.
Assessment and notification
- Each incident is assessed for severity and affected workspaces.
- For an actively exploited vulnerability or a severe incident we follow the EU reporting path: an early warning within 24 hours and a notification within 72 hours, and we inform affected customers.
- Personal data breaches are reported to you as the controller without undue delay, as the DPA sets out.
- Service status and incidents are published at status.kanman.ai.
After the incident
Severe incidents get a written review: what happened, why, and what changed so it does not happen again. Affected customers receive it on request.
Vulnerability reports
A person reads every report to [email protected], confirms it and keeps the reporter posted until the fix is out. See responsible disclosure.
Review
This policy is reviewed after every severe incident, and at least once a year.