Skip to main content
Gov AI

Structured analysis and classification-marked reporting

Analysis in Gov AI reads from the same governed entity store as collection, so the graph, the timeline and the map show the same subjects without re-keying. Structured analytic techniques are built in, from competing hypotheses to key assumptions and challenge reviews, and AI assessments are saved as reviewable records with their sources. Findings move through review to classification-marked reports and sealed evidence bundles.

See how it works
Link analysisStructured analytictechniquesAI assessments andfindingsTimelines and signalanalysisAuthorship andstylometryReports and evidencebundlesAnalysis andreportingPART OF THE GOV AI PLATFORM
Capabilities

What analysis and reporting gives your team

Link analysis

Explore relationships among people, organizations, accounts and places on an interactive graph with centrality measures, community detection, path finding and neighborhood isolation. An AI network assessment names the hubs, brokers and highest-risk actors for an analyst to review.

  • Degree, betweenness and PageRank centrality
  • Communities, paths and neighborhoods

Timelines and signal analysis

Swimlane timelines, density bands and a time scrubber establish sequence and bursts of activity. The signal lab detects anomalies and bursts across any collected series, projects them forward with prediction intervals and shows which signal tends to move first.

Structured analytic techniques

Hypothesis boards rank competing explanations by inconsistency and show which evidence is diagnostic. A key assumptions register arms each premise with falsifiers, challenge reviews let an independent colleague contest a product before release, and the scenario studio tests what-if assumptions on real baselines as labeled model output.

Authorship and stylometry

Compare writing style across accounts on phrasing, spelling, vocabulary, habits and rhythm, measured against a stated reference corpus and capped when a signal cannot be measured. Results support coordination and identity work as automated inferences, never as proof.

AI assessments and findings

Structured assessments state a threat level, narratives, entities, claims and recommended actions, and are saved as records rather than chat. Findings carry an assertion type, a confidence band, evidence for and against, and citations, and move through review and publication with versions. Horus, the page-aware copilot, answers questions in the context of the page an analyst has open.

Reports and evidence bundles

Produce classification-marked DOCX and PDF reports with charts, publish them to a library and distribute them by email to recipients cleared for the marking. Evidence bundles number and hash each exhibit, support redaction and a disclosure schedule, and seal the manifest so it can be verified later.

  • Classification markings
  • Clearance-checked distribution
  • Sealed, verifiable bundles
How it works

From collection to decision

  1. 01AssembleBring subjects from collection, a case or a search into the analysis workspace from the entity store.
  2. 02ConnectExplore links on the network graph and let centrality and community measures point to what matters.
  3. 03Place in timeLay events on timelines and maps to establish sequence, bursts, movement and hotspots.
  4. 04TestWeigh competing hypotheses, check key assumptions and put the product to a challenge review.
  5. 05AssessRun a structured AI assessment, review it and record findings with their assertion types and evidence.
  6. 06PublishIssue a classification-marked report or a sealed evidence bundle to cleared recipients, with every step audited.
In the field

Where it is used

FAQ

Questions buyers ask

Get started

Take the next step

See analysis and reporting working on your own sources, in a secure environment with clearance verification.

Explore Gov AI