Skip to main content

How AI Pet Matching Works

How it works

How AI Pet Matching Works

KCAI narrows thousands of possible comparisons down to a short list worth a human looking at. It does not decide that two animals are the same.

The four steps

Submit a report

Lost or found, plus the basic details of the animal.

Add photographs

Images and identifying marks carry the most weight.

KCAI compares records

Appearance, location, dates and description against compatible reports.

Possible matches surface

A person reviews anything promising before contact is made.

The signals that are compared

A match is built from several independent signals rather than one. When a photograph is unclear the other signals still carry weight; when a description is sparse KCAI does not pretend to more certainty than the evidence supports.

Appearance

Species, breed impression, primary and secondary colours, markings, size, coat length, sex.

Photographs

Visual similarity between submitted images. This measures whether two photographs look alike — not whether they are the same individual animal.

Location

How far apart the two reports are. A cat found three miles away is a very different proposition from one found thirty.

Timeline

Whether the dates make sense together. A pet cannot be found before it went missing.

Written detail

Collars, harnesses, injuries, behaviour and other identifying notes.

Confidence

A sparse report is not penalised for being sparse, but nor is it credited with certainty nobody has.

A person always reviews before anyone is contacted. Nothing reaches you automatically because software scored two reports highly.

What AI can and cannot determine →