Documentation Menu

SLA and Support

The value of a prod-ready platform is measured not only by model quality, but also by how support works during an outage. This page clarifies BriqMind's support level, response expectations, and incident communication framework.

01Incident Levels and Response Expectations

The support model defines not only communication channels, but also how critical an incident is considered and how quickly it will be handled. Incident levels must therefore be defined in a clear and reusable framework.

At the most critical level, the expectation is not only a fast reply, but active intervention and regular status sharing. Lower levels can be handled through planned release cadence or documented workarounds.

LevelDefinitionFirst responseTarget action
P1Full service outage or critical production impactResponse within 15 minutesActive intervention within 4 hours
P2Partial functional loss or significant performance degradationResponse within 1 hourAction plan within the same business day
P3Limited impact or error with workaroundResponse within 4 hoursPlanned release or documented solution

02Support Channels and Working Principles

A prod-ready support model is not complete by simply saying 'contact us'. It must be clear which channel is used for which incident type, who gets involved, and how often status updates are shared.

Support Channels

Email, ticketing system, dedicated customer channel, and live incident bridge for critical customers should be clearly defined.

Status Communication

Planned maintenance, outages, root-cause analysis, and next action steps should be shared in a standard format.

Example Incident Flow
P1 Incident Workflow

1. Customer notification or automatic alarm is triggered
2. Incident level is validated
3. First response is sent to the customer
4. Technical team starts active intervention
5. Periodic status updates are shared
6. RCA is prepared after the incident is resolved

03Operational Expectations

Maintenance Windows

For production customers, maintenance time, potentially affected services, and rollback conditions must be defined in advance.

Root-Cause Analysis

After critical incidents, not only the fix but also RCA and a permanent prevention plan should be delivered to the customer.

Status Updates

Especially for P1 and P2 incidents, a regular information cadence must be maintained until resolution is complete.

A strong SLA does not only promise a fast response; it also clearly defines maintenance, notification, escalation, and RCA behavior. This is where enterprise customer trust is formed.