Content safety combines ingestion guardrails, whitelists, and bot-level pauses.

• Document gating: uploads require terms acceptance and pass antivirus/safety checks.
• User gating: whitelist testers before opening access broadly.
• Runtime controls: pause bots, reset flags, and fine-tune personas.

Stay compliant while uploading Documents.

• Confirm you own/have rights to the documents; reject prohibited content.
• Use Upload & ingest flow to track statuses (waiting, processing, ready, failed).
• Retry failed documents after fixing issues; watermarks are available per tenant.

Limit who can message bots during pilots.

• Manage bot user lists from the WhatsApp tab.
• Enable/disable whitelisting; add/remove users inline.
• Keep a small allowlist when testing sensitive knowledge bases.

Use the operational cockpit to keep bots healthy.

• Pause/resume bots if behavior drifts.
• Adjust purpose and docs summary to tighten scope.
• Monitor live transcripts and logs where enabled.

Handle bad interactions gracefully.

• Reset flags on users after reviewing behavior.
• Use message handling preferences (tone, disclaimers) to align with policy.
• Escalate risky chats to human review as needed.

Jump straight to the controls inside the app.