Service level agreement
What happens if HansaChat goes down.
Availability and recovery objectives, published alongside external service measurements.
Updated: 11 September 2026
What this agreement is
Operational objectives for availability, incident communication, support response and data recovery for paid workspaces, unless an individual order agrees otherwise. For free and demo workspaces these figures are guidance rather than agreed service levels. The status page shows live measurements and recorded incidents. The Terms of Service govern liability.
See live status and history- Availability target
- 99.9% monthly
- Measured by an independent external monitor every 5 minutes.
- Recovery point objective
- 6 hours
- Target maximum data-loss window following a failure.
- Recovery time objective
- 24 hours
- Target time to restore the core chat service.
- Restore testing
- Quarterly
- Isolated restores, plus an annual end-to-end exercise.
Availability
A 99.9% monthly availability target, measured externally.
HansaChat aims for 99.9% monthly availability of the chat service. Availability is the share of five-minute checks that succeed, measured by an independent monitor on Cloudflare — not by HansaChat itself. The full measurement history is public.
What the measurement covers
- Chat API, database, cache, message queue, and realtime connectivity
- The public website (hansa.chat)
- Every five minutes, from outside HansaChat’s infrastructure
Not counted toward downtime
- Maintenance that is announced on the status page in advance
- Problems isolated to a customer’s own network, client device, or integrations
- Force majeure events
What counts as downtime
A five-minute check counts as downtime when the chat health endpoint cannot be reached or reports one of the core services as unavailable. Monthly availability is calculated per calendar month from these checks.
Incident communication
Incidents are detected and published automatically.
The monitor opens an incident after three consecutive failed checks (15 minutes) and resolves it with the first successful check. The public status page reflects incidents without manual publishing, so it stays truthful even at 3 a.m.
Post-mortems for significant incidents
For significant incidents, HansaChat publishes a post-mortem on the status page describing cause, impact, duration, and corrective action.
Support response
First-response times you can plan around.
Support reaches HansaChat at igor@hansa.chat and through the in-product contact route. First-response targets during business hours:
| Severity | Definition | First response |
|---|---|---|
| Critical | The chat service is unavailable for all workspace members. | Within 4 business hours |
| High | A core feature is unusable with no workaround. | Within 1 business day |
| Normal | Questions, minor issues, and configuration help. | Within 2 business days |
Business hours are Monday to Friday, 9:00–18:00 CET/CEST. Reports received outside business hours count from the next business day. The service itself is monitored around the clock by the automated external checks. Public holidays at the operator’s location in Lübeck are excluded. First response means an initial human response, not resolution; automated monitoring does not promise staffed support outside these hours.
Data recovery
Documented recovery objectives, tested on a schedule.
If a severe failure damages or destroys service data, recovery follows documented objectives rather than improvisation.
| Objective | Target | What it means |
|---|---|---|
| RPO | 6 hours | The target maximum data-loss window, derived from the backup cadence; actual recovery depends on the incident and available backups. |
| RTO | 24 hours | The target time to restore the core chat service and workspace data after a severe failure. |
| Supporting services | 48 hours | Website, billing, provisioning, and supporting services target restoration within 48 hours. |
Recovery basis
Recovery builds on IONOS Cloud MariaDB Database as a Service with seven-day point-in-time recovery and HansaChat’s encrypted application backups stored in Frankfurt and Stockholm. The Security page documents the backup locations and retention schedule.
Review backup details on the Security pageRestore testing
Backups only count if they restore. HansaChat commits to an isolated restore of workspace data every quarter and an end-to-end recovery exercise at least once a year, with results recorded internally.
Scope & boundaries
What this agreement does not promise.
- Availability, first-response and recovery figures are operational targets, not guarantees of a particular outcome. Missing a target does not, by itself, create an entitlement to automatic compensation or service credits. Liability, including the subscription-fee cap for business customers and its exceptions, follows section 8 of the Terms of Service. Mandatory statutory rights remain unaffected.
- Nothing in this agreement is a security certification. HansaChat documents its implemented controls on the Security page.
- A dated website update alone does not change an existing agreement. Changes to agreed service levels follow the notice and agreement rules in the Terms of Service.