RootInference

Observability

Every request, token and credit accounted for — across all providers in one place

Unified request logs, latency and cost dashboards, and spend attribution by key, team and feature — across every provider you use. Set budget alerts, export the ledger, and answer 'what did we spend on inference and why' in seconds instead of spreadsheet archaeology.

Explore the dashboardsIncluded on all plans
Tracerootinference usage --by team --last 30d
  1. 30-day totals: 4.2M requests, 9.8B tokens, 41,205 credits

  2. search-team used 18,900 credits, 71% on economy-tier models

  3. agents-team used 14,050 credits at p95 latency 2.1 s

  4. Alert raised: copilot key at 84% of monthly budget

  5. Ledger exported to usage-2026-08.csv

  6. Credits expiring in next 90 days: 0 of 62,300

Capabilities

What it does

Turn opaque multi-vendor AI bills into attributed, alertable, exportable spend data.

  • Full request logs with model, provider, latency, tokens and cost
  • Latency and cost dashboards across all providers, unified
  • Per-key, per-team and per-feature spend attribution
  • Budget alerts and hard spend caps per key or organisation
  • Exportable credit ledger for finance and chargeback
  • Model and provider comparison reports from your own traffic
  • Credit balance and expiry tracking — every credit valid 12 months

See Usage Observability on your own traffic

Point your existing OpenAI SDK at our base URL and try it in minutes — or book a demo to plan a production rollout.