This feature is currently in private beta. Contact us to join.
Availability
Private beta
Contact us to join.
- Create entity types and hierarchies
- Hard limits for credits and metered features
- Visibility and sorting across all entities and cardinalities
Early access
Contact us to join.
- Allocated budgets per entity
- Soft limits
- Alerts
The problem
Enterprise customers adopting AI features face a critical governance challenge: uncontrolled AI usage and spend. Without proper controls:- Individual users or agents can exhaust shared credit pools
- Finance teams lack visibility into departmental consumption
- IT admins cannot enforce cost boundaries across teams
- Usage spikes cause unexpected billing surprises
What you can build
How it works end to end
- You define the entity types once per environment — for example
org,team,user,agent. - Your backend provisions entity instances for each customer as they are created in your product (their org structure, team members, AI agents). This is a vendor-side operation, not customer-driven.
- Your customers set the actual budgets and limits through an in-app interface you build.
- Your application calls Check before any governed action and Ingest to record consumption.
Governance is opt-in
If an entity has no assignment, the Check endpoint grants access. You can roll out governance incrementally — ungoverned entities continue to work exactly as before.Next steps
Core concepts
Understand entity types, entities, assignments, and the hierarchy model
Setting up governance
Step-by-step guide to configuring entity types, entities, and assignments
Check and ingest
Integrate the runtime check and ingest endpoints into your application
Query the governance tree
Fetch usage and budget data for dashboards and admin UIs