Command brings together agent analytics, usage and cost reporting, conversation feedback, Knowledge reports, document reports, and Loop run reports.
In this section
- Pull agent analytics for the right organization and region.
- Review monthly usage by model, effort, agent, and day, then export the result.
- Use Project Reports to track conversion, training, and processing progress.
- Open Document Reports for content, AI output, and notes.
What's inside
- 01
Reports page
Sidebar → Reports. Pick an organization and an agent to see that agent's analytics. Stats come from the matching region (US or EU) based on the agent's region setting.
- 02
Conversation feedback
Feedback on AI conversations is org-scoped through the agent. Status is derived from timestamp and error fields: Failed (review errored), Completed (review finished or timestamp recorded), Running (started, not finished), Pending (queued). Training → Human Feedback is where you sync, review, and annotate individual conversations.
- 03
Usage and cost reporting
The Usage page shows monthly token and cost totals with breakdowns by model, reasoning effort, and agent, plus a daily usage chart. CSV export provides the same detail for analysis. Some older or externally reported activity may not include every attribution field, so the page calls out incomplete model or agent attribution rather than assigning it incorrectly.
- 04
Project and document reports
Every project has a Project Report summarizing metadata and AI progress across Conversion, Training, and Processing — including a PII Scan card with Files Scanned and Files Masked. Every document has a Document Report with its content, AI output, and notes, exportable to PDF, DOCX, or Excel.