Architecture

CaseTrace separates computation, notification, and duty. The system may compute similarity and surface information immediately. Use of that information remains human-initiated.

Architecture overview

Sources → Controlled ingestion → Event structuring → Preserved records → Human review → Export

Separation model

CaseTrace separates key concerns to prevent conflation of computation and authority.

  • Data ingestion from interpretation
  • Computation from conclusion
  • Notification from obligation
  • Infrastructure from authority

Data ingestion is limited to configured sources; ingestion does not interpret intent. Computation (for example similarity or correlation) produces candidate associations; computation is not a finding. Notifications surface availability of data; they do not carry enforcement or escalation authority. Infrastructure provides audit, provenance, and access controls; those controls do not assign investigative authority.

Controlled ingestion

Inputs are explicitly configured. The system does not infer new sources.

Similarity and correlation

CaseTrace may compute similarity instantly, including possible feature, object, license plate, and cross-location correlations.

These results are similarity-based and non-conclusive.

Informational notifications

When new records or possible correlations are available, logged-in users may receive in-application notifications.

Notifications indicate availability, not urgency.

Human review

Review occurs only when an authorized user affirmatively chooses to investigate. No timeline is imposed by the system.

Export and packaging

Packaging is deliberate and attributable. The system does not escalate on its own.