How AskVault handles incidents and outages
What to expect today
We are a small team, so here is the honest version. We watch for errors and health-check failures, act on them as soon as someone sees them, and tell affected customers by email. We do not run a staffed round-the-clock rotation, we do not promise fixed response times on self-serve plans, and there is no public status page yet. Enterprise customers can negotiate response terms in their contract.
How incidents are found
- Monitoring. Error tracking and the API health endpoint surface failures to the engineering team.
- Customer reports to [email protected].
- Security reports to [email protected].
- Provider notices from the services AskVault runs on.
How severity is judged
- P0, full outage. The agent cannot answer on any channel.
- P1, partial outage. Some customers or channels are affected.
- P2, degradation. The service works but is slow or a feature is impaired.
- P3, minor. Edge-case bugs and non-critical dashboard problems.
Severity decides who is pulled in and whether customers are emailed.
How we contain an incident
Containment comes before root cause. The usual tools:
- Switch off a feature. Many changes ship behind a feature flag that we can turn off without shipping new code.
- Roll back a release to the last known-good version.
- Restart or restore. Restart the affected service, or restore the database with the provider's point-in-time recovery. AskVault runs in a single region in the United States, so there is no standby region to fail over to.
- Tighten rate limits when one source of traffic is causing the problem.
How you hear about it
- Email to workspace owners for incidents that affect them, at the start, at major milestones and at resolution.
- Incidents affecting customer data are notified to affected customers without undue delay once confirmed, as a data processor must under GDPR. See GDPR.
- A written summary of a significant incident is available to affected customers on request from [email protected].
Reporting a problem
- Service problems: [email protected].
- Security problems: [email protected]. There is no bug bounty; we will credit you once a fix ships if you would like.
FAQ
Is there a public status page?
Not yet. See status page for how to get incident status today.
Do you commit to response times?
Not on self-serve plans. See SLA.
What about incidents at your providers?
When a provider's incident affects AskVault customers, we email affected workspace owners and say which provider was involved.