CCTV evidence for insurance claims, without the scrubbing

A claim arrives weeks or months after the event. Somebody exports a day of footage from four cameras, and someone else, on one side of the claim or the other, has to watch it. The moment that decides the claim is ninety seconds long, nobody knows which ninety seconds, and so the review costs an afternoon per camera and things get missed. The stakes aren't small either: the average slip-and-fall claim settles above $20,000 (National Floor Safety Institute), and timestamped footage of the hazard, or its absence, decides most of them.

From hours of footage to a timeline

Two report checks are built for exactly this. Both read a whole video rather than judging frames as they arrive, and both produce documents rather than alerts:

• Incident Reconstruction writes the timeline of an event: who entered frame when, what happened in what order, with the timestamps to check against the claim.
• Liability Evidence Pack assembles the relevant clips and moments into a bundle you can hand to an insurer, an adjuster or a solicitor, instead of a hard drive of raw footage.

Several cameras are handled by the Library rather than by tracking anybody. Drop each camera's export into one incident folder (folders nest, and hold recorded and live footage alike), then run a check across the whole folder in one action. Underneath the reports sits the search every account gets: type “someone slipped in the entrance” and the moment is found in seconds, from every clip in the folder at once.

What it costs

The price comes from your own usage rather than a rate card, with a $9 a month minimum, and the whole derivation is shown on screen at signup before you pay. Report checks sit at the cheap end of the Store; each is priced as the extra it would add at your usage, from a $3 a month floor. A business that keeps Svid running year-round has searchable footage on the day the claim letter arrives.

For insurers and lawyers reviewing footage

When the footage lands on your desk, the same tools work in reverse. Load the exported files (Svid reads uploads and video URLs as happily as live feeds), search for what the claimant describes, and check whether the reconstructed timeline supports it. The review that took an afternoon takes about as long as typing the sentence.

Getting started

Create an account at app.svid.ai, answer the four onboarding questions, and pay by card with the arithmetic in front of you. Most claims work starts with exported clips from an incident that's already happened: make a folder, upload the files, run the report checks, and start asking questions.

Can it work on footage exported after the incident, or only live cameras?

Exported footage is fine, and it's the usual starting point for claims work. The pipeline treats live and recorded footage identically; upload the files or point it at a URL and everything becomes searchable the same way.

Will the output be accepted by my insurer or in court?

That's your insurer's or your lawyer's call, and we won't pretend otherwise. What Svid produces is built from your normal CCTV exports; its job is to find and organise the relevant clips so the people who decide can see what matters.

How do I find one person in the footage without facial recognition?

By describing them. Search reads the plain-English descriptions of every frame, so “the man in the high-vis jacket near the loading door” finds the matching moments in whatever folder you're searching. Nothing follows a person automatically and nothing biometric is processed; you review the moments and draw the connections yourself.

What does it cost if I only need it for one incident?

It's a monthly subscription with a $9 minimum, priced from your own usage, and the arithmetic is on screen before you pay. The report checks are priced from your usage too, from a $3 a month floor, so working up a single incident on exported footage is a cheap way to find out what it actually shows.