For veterinarians and behaviourists
The history the owner could not quite describe.
A consultation usually starts with a recollection. This gives you dated observations, the context around each one, and a measurable comparison against how the animal normally behaves.
What the work looks like
Invite the owner to share a profile
The owner keeps ownership of the data and can revoke access at any time. You see the behavioural record, not their whole account.
Look at deviation, not absolute values
A cat that vocalises a lot is not a finding. A cat vocalising three standard deviations above its own six-month baseline might be.
Add expert annotations
Mark a prediction as clinically irrelevant, uncertain or worth monitoring. Your annotation sits alongside the model output permanently and is never overwritten by it.
Generate an observation report
Dated, with confidence, evidence, data completeness and the caveats — in a form you can put in a record or hand to a colleague.
What you get
Longitudinal behavioural change per individual
Baseline deviation with sample counts, so thin baselines are visible as thin
Flagged anomalies and possible pain indicators as observational flags
Expert annotation that outranks and outlives model output
Clinical observation reports in PDF, CSV or JSON
Client and case management
Explicit not-a-diagnosis framing on every relevant output
Observational flags, not findings
Pain and discomfort outputs are raised when measured signals overlap with those associated with discomfort in the literature. They are not clinical findings, they name no condition, and they suggest no treatment. The product states this on every such result, to the owner as well as to you.