Atomation-hosted
Fastest path for most customers. Atomation hosts the portal and data plane, with tenant isolation, encrypted credentials, immutable snapshots, and controlled access.
Deployment
Your Okta data stays in a boundary you control.
Deployment should match the sensitivity of the environment. Atomation can start simple or move the data plane closer to the customer-approved or partner-approved boundary when needed.
Okta Identity Engine only Optional AI summaries Hosted or self-hosted capable
Data models
Three ways to run the assessment.
Customer or partner self-hosted
Run the data plane in a Docker-based environment you or a partner controls. This is the right pattern when identity data must remain inside your boundary.
Hybrid control plane and data plane
Keep sensitive snapshot and report processing in your environment while centralizing configuration, update delivery, and operator workflows where appropriate.
AI boundary
The executive summary does not require raw identity data leaving your environment.
Atomation's narrative layer is designed around findings, not raw user exports. The AI provider is pluggable so customers can use a local model, an enterprise-approved model, or a partner-controlled deployment pattern.
What stays controlled
- Okta snapshots and findings remain in the selected deployment boundary.
- AI receives a controlled findings summary, not raw identity data.
- Reports are frozen artifacts so exports remain consistent over time.
Get started
Need the Okta assessment to run inside a specific boundary?
Book a free discovery call. We'll map your workflow, find the highest-value automation, and show you the smallest useful first build — no obligation.