Usage dashboard
Investigator-level AI consumption: calls and credits by model, by service bucket and by case, with alerts when a tier gate or credit cap is hit. Accounting only — usage is not evidence about any person.
Cap & tier alerts
Raised from live wallet headroom and refusals already recorded by the gateway.
No tier or credit-cap issues in this window. All buckets have headroom.
Case budgets
Set a credit spend cap per investigation. Spend is counted across all recorded calls for that case, and breaches surface in the alerts above.
No case budgets set. Pick a case above to cap its AI spend.
By model
Which models the investigation work actually used.
No model calls in this window.
By service bucket
Which metered wallet each call drew from.
No service usage in this window.
By case
Per-investigation spend. Calls made outside a selected case are grouped as unattributed.
No case-attributed calls in this window.