Overview
Adaptive AI enables you to automate the monitoring and maintenance of your AI models. By configuring custom rules based on observability data, the tool detects anomalies and generates actionable alerts. Unlike standard notifications, these alerts allow you to apply pre-configured remediation actions directly from the interface.
Getting Started
Ready to set up your first rule? Start with the Step-by-Step Tutorial.
Core Components
The Adaptive AI workflow operates across three main interfaces:
-
Main Dashboard
The central hub containing a high-level summary of model health and a prioritized list of active alerts. Use this to get an "at-a-glance" view of your ecosystem.
-
Rules Configuration
The setup engine where you define the logic for specific models. Here, you configure:
-
Conditions: The specific triggers (Metrics-based or Temporal) that define an event.
-
Actions: The remedial steps (e.g., Rebuild, Replace) available to resolve the alert.
-
-
Monitor and Adapt
The operational page for reviewing the full history of alerts for a specific rule. This is where you inspect the details of an active alert and apply the selected remediation action.