Security footage becomes more useful when you can investigate it.

From footage to a finished investigation

Six stages, each designed to keep the underlying evidence visible and the human investigator in control.

1. 01 / Ingest

   Start with footage from connected live cameras, existing NVR/VMS systems, or footage uploaded after an incident. VIGIL processes it and prepares searchable observations for investigation.

2. 02 / Find

   Start with what you know — a location, a time window, a description, an incident. VIGIL surfaces relevant footage so investigators spend less time searching and more time understanding what happened.

   • Time
   • Location
   • Camera
   • Visible characteristics
   • Known observations
   • Incident context

3. 03 / Follow

   Move from an initial sighting to related footage while keeping the underlying video and timestamps accessible. When continuity cannot be established reliably, it stays unresolved for review.

4. 04 / Trace

   Review where a subject was observed, when they were last confirmed, relevant gaps in visibility, and subsequent candidate observations — with the line between fact and gap always visible.

   • First confirmed observation
   • Last confirmed observation
   • Camera observations
   • Visibility gaps
   • Unresolved periods

5. 05 / Ask

   Ask questions about the available evidence and navigate the investigation conversationally, instead of manually searching every recording one camera at a time.

   • Show me activity around this incident
   • Where was this subject last observed?
   • What happened before this event?

6. 06 / Review

   Computer vision is powerful, but security footage can be ambiguous. Uncertain identity and continuity decisions remain reviewable rather than being silently forced into a match.

   • People cross paths
   • Subject becomes occluded
   • Tracking interrupted
   • Video quality poor
   • Evidence conflicts