Current combined platform and AI operating spend
Run AI and platform services with the same financial accountability as the rest of IT.
The FinOps surface is now route-native in `apps/web`, bringing AI usage, observability, cloud services, and vendor-backed platform spending into the same production shell as the rest of the operating model.
Estimated savings from lower-tier AI routing discipline
AI demand using high-cost tiers this month
Open budget or cost-governance exceptions
Cost Domains
FinOps should connect spend to the service areas that create it and feed an owned optimization backlog.
| Cost Domain | Current Trend | Main Driver | Owner / Action |
|---|---|---|---|
| AI routine tiers | Stable and within forecast | Knowledge and request-intake volume | Platform admin / keep low-risk routing at Tier 1-2 |
| AI premium tiers | Rising slightly | Incident command support and leadership analysis | IT leadership / review premium-use approvals |
| Observability tooling | Healthy and predictable | Metric retention and synthetic-check growth | Infrastructure lead / review data-retention policy |
| Remote access vendor service | Watch | Escalation surge and premium support-hours use | Vendor owner / challenge overage path |
| Web and data platform services | Within target | Planned publishing and scheduled refresh workloads | Service owners / no action required |
Current Cost Governance Levers
These are the operational decisions that most directly affect spend without damaging service quality.
Route routine AI down
Keep standard knowledge and intake demand on lower-cost tiers unless risk or ambiguity requires escalation.
Control premium AI by approval
Require explicit justification before specialist or premium tiers are used for leadership or incident work.
Challenge vendor overages
Use escalation evidence and volume patterns to dispute or renegotiate surge support charges.
Tune telemetry retention
Keep observability useful without accumulating high-cost storage or duplicate data streams.