Skip to main content

Data handling

No model training on your data

Wild Moose does not fine-tune or train AI models on your data. Instead, we analyze patterns across your systems to generate human-readable investigation plans that help resolve incidents faster.
Wild Moose does not store logs, metrics, or other raw data. All processing happens in-memory and is ephemeral. The only retained output is structured metadata in the form of investigation plans, which can be reviewed to ensure no sensitive content is included.
All integrations operate strictly in read-only mode. Wild Moose queries your observability tools during investigations but never modifies your dashboards, alerts, or infrastructure configurations.
Each integration uses the minimum permissions required. For example, Slack access is limited to channels you explicitly configure - Wild Moose does not read messages in other channels.

Data usage

Wild Moose uses AI to analyze and correlate data across your environment for two primary purposes:
  • Infer investigation flow - based on investigation conversations, observability audit trails, and documentation, Wild Moose identifies typical troubleshooting patterns and workflows.
  • Find root causes in real time - using logs and metrics, Wild Moose detects patterns and relationships to expedite issue resolution.

Compliance & encryption

  • SOC 2 Type 2SOC 2 Type 2 compliant - Wild Moose adheres to SOC 2 Type 2 standards for data security, availability, and confidentiality.
  • End-to-end encryption - all communications and data exchanges are encrypted to ensure full confidentiality.
  • Secure credential storage - API tokens and integration credentials are stored in a secrets manager and encrypted at rest.

Access requirements

To enable Wild Moose for your environment, we require read-only access to the following:
CategoryWhat’s needed
Logs & metricsA read-only service account and API credentials for your observability platform (Datadog, Grafana, Splunk, etc.)
Runbooks & documentation (optional)Either read-only API access to relevant sources, or a one-time export of relevant information
Code & deployments (optional)Code repositories or deployment tools can be configured based on your stack
All integrations are opt-in. You can connect as few or as many data sources as you’re comfortable with - Wild Moose works with whatever you provide.
Have security questions? Contact us at [email protected].