You point it at an app's API documentation and name the access you want to govern. An AI research agent reads the docs live, works out how the API authenticates and returns data, maps it to Oleria's identity model, and assembles a connector manifest. You confirm each stage, and Oleria validates the live connection before anything goes live.

No. Integration Studio produces a connector manifest, not code — a structured description of the app's auth, endpoints, and field mappings that a generic runtime reads and runs. There's no engineering cycle and nothing to maintain in a repo.

The same things Oleria's hand-built connectors do: read users, groups, and entitlements into your identity model, and act on them — disable an account, remove a group membership, or run an access-review campaign, all without leaving Oleria.

Two checks. Every draft is checked by a deterministic validator that runs entirely outside the AI model — a mechanical gate that can't be talked around. And you confirm each stage yourself before the agent moves to the next one, with full visibility into what it found and where.

Oleria AI answers questions about the access and risk already in your environment. Integration Studio is what gets a new app into that environment in the first place — it builds the connector that makes an app visible to Oleria at all.

A link to the app's API documentation, and the credentials Integration Studio will use once you approve the connector. Nothing else — no engineering time, no vendor roadmap request.