Skip to main content
Track and audit all user actions and system events in AI for Process, including logins, role changes, model updates, and workflow activities. Use audit logs to identify patterns in user activity, detect anomalies, spot unauthorized usage, and ensure compliance with internal policies and regulations. Navigation: Go to Settings> Monitoring > Audit Logs.

Audit Log Dashboard

Each audit log entry captures the following information: User ID, Role ID, Model ID, Workflow ID, Guardrail ID, Integration ID, and Experiment ID are the unique identifiers for the respective module entities in the system.
IP Address is fetched from the user’s current network.
Audit logs dashboard

Filter Audit Logs

Select the Filter icon to narrow down logs by category, event, or user using Is Equal To or Is Not Equal To operators. Combine multiple filters using AND (all conditions must be met) or OR (any one condition must be met) — operators cannot be mixed in the same filter set.

Export Audit Logs

Select the Export icon on the Audit Logs page to download a CSV file containing all audit log entries, including date and time, event name, category, description, user or app name, and IP address.

Audit log event reference

The tables below list all tracked events by category.

Account-level events

The User ID and IP Address are shown for audit log entries across all modules, in addition to module and category-specific metadata listed below.
Metadata: Email ID is shown for all events in this category.
Metadata: Role ID is shown for all events except Role Changed.
Metadata: IP Address and User ID are shown for all events.
Metadata: App ID, IP Address, and User ID are shown for all events.
Metadata: IP Address, Integration Name, Integration ID, Integration Type, and User ID are shown for all events.
Metadata:
  • IP Address, User ID, Model ID, and Model Name are shown for all Model events.
  • Model Type is also shown for: Model Added, Model Deleted, API Key Created, API Key Deleted, Model Fine-tuning, Model Deployed, and Model Undeployed.
  • Hardware Type is shown for: Model Fine-tuning, Model Deployed, and Model Undeployed.
Metadata: Workflow ID and Workflow Name are shown for all events.
Metadata: User ID, IP Address, Agent ID, Node name, Node ID, and Node Type are shown for all events.
Metadata:
  • Prompt ID is shown for all Prompts events.
  • Prompt Name is shown for: Prompt Created, Versions Committed, Versions Restored, Prompt Shared, Endpoint Copied, and API Key Created.
Metadata: User ID, User Name, IP Address, and Custom Script Name are shown for all events.
Metadata: Guardrail Name, Guardrail ID, and Hardware Type are shown for all events.
Metadata: User ID, User Name, IP Address, Agent ID, Node name, Node ID, and Node Type are shown for all events.
Metadata: User ID, User Name (where applicable), IP Address, Connection Name, Connection ID, Deployment Name, and Hardware Type are shown for all events.

Workflow-level events

User ID, IP Address, and Workflow ID are included in all workflow-level entries, in addition to any category-specific metadata listed.
Metadata for deployment events: Workflow version, API mode, sync/async, and URL are shown for Workflow Deployed and Workflow Undeployed events.Metadata for version and configuration events: Version ID is shown for the events below.
Metadata: User ID, IP Address, Agent ID, Node name, Node ID, and Node Type are shown for all events.