Settings

Configure your organization, agent behavior, and security posture.

AI agent controls

Enable AI agent

Allow OpsPilot to run workflows and draft outputs on behalf of users.

Require human approval before external sharing

Outputs cannot be shared outside the org without an approver sign-off.

Allow third-party LLM processing

Permit calls to hosted models like OpenAI / Anthropic. Disable to restrict to private models only.

Use only private / self-hosted models

Restrict the agent to your self-hosted LLM gateway.

Store prompts and outputs

Retain prompts and generated outputs in your tenant for audit and re-use.

Retention period

How long prompts, outputs, and run history are retained.

Data residency

Region where prompts, outputs, and indexes are stored.

Never train on customer data
enforced

No customer data is ever used to train base models.

Show citations in every output
enforced

Every agent-generated output must include citations to source evidence.