The agent prepares. The clinician decides. Everything is on the record.
Nuravenn is built so automation never makes a clinical or readiness determination. It does the assembly and the math; the clinician holds the judgment and the signature.
The clinician is the final authority, by design.
Nuravenn drafts the summary, the scores, and the recommended readiness category. It never assigns readiness, never signs, and never closes a case on its own. Judgment-only decisions stay with the clinician.
The determination is computed, not generated
A deterministic rules engine produces the readiness disposition; language models write the narrative around it. The same inputs give the same disposition every time, and the wording can be regenerated without the decision moving. This is a stronger guarantee than an explainable model, because there is no model in the decision.
Runs inside the governing instructions
Assessments execute against the applicable health instructions as machine-checkable rules, and every rule carries the authority it derives from, so a flag can be traced to the instruction behind it rather than to a developer's judgment. When policy changes, it is configured low-code, not re-engineered.
Every output is traceable
Each flag, score, and recommendation carries its source data, the rule applied, and the person who signed. Every agent run and every model call is recorded with its trigger and its output, so the trail covers the automation as well as the clinical decision. Audit-ready by construction, not after the fact.
Access follows the real sign-off chain
Roles mirror how an assessment is actually cleared rather than a generic administrator and user split: service member, record reviewer, behavioral-health provider, certifying provider, administrator. Permissions are enforced at the interface between the application and the data, not merely hidden in the screen.
Designed to IL5, FedRAMP & HIPAA standards
CUI-aware data handling, role- and attribute-based access, and deployment inside your boundary. Data stays where it lives; Nuravenn overlays it.
One assessment, every service
The same governed layer runs the assessment across the services and defense health divisions — active component or reserve, in an MTF or through a readiness contractor — so a capability proven in one program extends enterprise-wide without a migration.