Seven stages from approved evidence to reviewable working output.

Authorized users upload approved non-CUI, native-text evidence. Qualense checks supported file constraints, integrity, malware results, secret patterns, and apparent CUI markings. Images and scanned/image-only PDFs are currently quarantined before external AI processing. These checks supplement — but do not replace — customer classification and authorization.

The AI-assisted pipeline proposes mappings between evidence excerpts and assessment objectives. Reviewers must inspect the source, exact excerpt, applicability, and provenance before accepting a suggestion.

A separate verification operation evaluates whether a selected passage supports a proposed mapping. Its result is decision support and does not replace human review.

Nine deterministic rule families run alongside the review workflow for conditions such as undated or stale evidence, screenshot-only support, draft material, log retention, policy without practice, method mismatch, and scope. Their results are prompts for qualified review.

Assessment records expose readiness and evidence-effect states, including gaps and items needing review. Qualense distinguishes insufficient or contextual material from evidence that supports implementation.

Approved assessment state can be recalculated into a supporting result on the −203 to +110 scale with requirement-level attribution. The customer validates the catalog, method, inputs, and total before any submission.

Generate a gap report DOCX, POA&M XLSX, supporting SPRS summary XLSX, evidence index XLSX, or a ZIP containing those four. Supported branding includes firm name, logo, primary color, and footer.