Analytics — Usage Insights

Navigate to: Dashboard → Usage Insights

What it shows: Per-employee and per-team breakdown of API usage, cost, and token consumption. Designed for administrators who need to identify top consumers and allocate costs.

Key panels:

Panel	Description
Total Requests	Org-wide API calls in the selected period
Total Tokens	Combined input + output tokens
Total Cost	Estimated spend based on provider pricing
Requests by Employee	Ranked list of employees by request volume
Cost by Model	Spend breakdown per LLM model
Active Employees	Count of employees who made at least one request

Step-by-step:

1. Click Usage Insights in the sidebar.
2. Set the date range using the picker at the top.
3. Click an employee row to expand their request history for the period.
4. Click a model row in Cost by Model to filter the employee table to only requests using that model.
5. Click Export CSV to download the full breakdown for finance reporting.

How to test:

1. Send requests using two different employee API keys (create keys in Employee Management if needed).
2. Reload Usage Insights — both employees should appear with their respective request counts.
3. Verify cost estimates reflect the model’s published pricing.

API endpoints:

• GET /admin/usage/stats — aggregate stats
• GET /admin/budget/report — cost breakdown per tenant/model

---