> For the complete documentation index, see [llms.txt](https://intelfactor.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://intelfactor.gitbook.io/docs/intelfactor-ai/validation-and-rollout/validation.md).

# Validation

How IntelFactor is validated on-site, and what evidence is produced.

Defines on-site validation gates and the evidence you should expect.

### What this covers

Validation means proving the system behaves as specified in your environment.

This documentation stays customer-safe.

It focuses on:

* Acceptance criteria and rollout gates
* Evidence and traceability expectations
* Operational checks before and after go-live

### Typical rollout gates

Teams usually validate in stages.

* **Connectivity gate** Event delivery works end-to-end.
* **Continuity gate** Outcomes continue locally during WAN loss.
* **Evidence gate** Each outcome links to auditable evidence artifacts.
* **Review workflow gate** Humans can review, override, and annotate outcomes.

### What “acceptance tests” look like (conceptually)

Acceptance tests should be observable and repeatable.

Examples:

* Defect injection or representative defect sets
* Known-good control samples
* Latency and throughput checks at the inspection point
* Failure drills (power loss, network loss, storage pressure)

### Out of scope

We do not publish:

* Model internals or training methods
* Customer thresholds and tuning values
* Site-specific runbooks

For deployment-specific validation plans, contact your IntelFactor point of contact.

### Next steps

* [Acceptance tests & rollout gates](/docs/intelfactor-ai/validation-and-rollout/acceptance-tests-and-rollout-gates.md)
* [Evidence & retention policy](/docs/intelfactor-ai/architecture-and-data/evidence-and-retention-policy.md)
* [Interfaces](/docs/intelfactor-ai/architecture-and-data/interfaces.md)