Context for a veterinarian, not a substitute for one.
WhiskerMatch is not veterinary software. It does not diagnose, prescribe, interpret clinical findings, assess risk, or set treatment. What it does is keep the medical context of an animal's life in the shelter system dated, attributed, and routed to the right human, so veterinary judgment lands on a complete picture instead of a fragment.
What the record actually holds.
Four kinds of rows, all with the same discipline: dated, attributed, sourced. The system organizes them and routes their follow-ups. Meaning stays with people.
Medical events
An event carries its type, the date it happened, the provider when there was one, the observation in the reporter's own words, treatment if one was given, and a follow-up date when one was set. Every row records who entered it and when. Nothing arrives without a name attached.
Vaccinations and medications
One row per vaccination, with a due date and an administered date; medications and weight records sit beside them. Due dates that pass without completion surface on the follow-up queue, read from stored dates and states, never scored.
Veterinarian sign-off
A medical event can carry the veterinarian it belongs to, and a separate signature when a veterinarian signs it. Attribution is the point: the record knows whose professional judgment stands behind a row, and which rows have none.
Behavior observations
A behavior observation requires context and the observation itself, both in the reporter's words. Severity is a required choice the human makes, low, moderate, high or urgent, with no default and no inference. Review is recorded once and stays immutable; marking an observation as affecting the public profile requires a written note explaining why.
The through-line is attribution. A medical row without a source is a rumor; a medical row with a source, a date, and a signature is context a veterinarian can work with. That is the whole of what the software contributes to veterinary medicine: it refuses to let the context go missing.
An animal with an unresolved medical question does not publish.
The publish gate is a hard check, not a suggestion. Before a public profile can go live, the system reads every active hold on the animal. Any unreleased medical hold blocks publishing, with the blocking categories named in the error. Released holds never block.
The three medical hold categories
Urgent veterinary
Something needs a veterinarian's attention before the animal is listed.
Unresolved medical source
A medical note exists that nobody has finished interpreting or confirming.
Missing medication instruction
A medication is on the record but its instructions are not complete.
Legal, bite-quarantine, stray and behavior-safety holds are real stops in their own flows, but they are not medical holds and do not pass through this gate. Each kind of caution has its own mechanism, and the mechanisms are not mixed.
Behavior notes are observations, never verdicts.
A behavior note is not a sentence. The system is deliberately built so that nobody can quietly turn an observation into a label.
Severity is chosen by a human, never inferred
Severity is chosen by a human, never inferred
Severity is a required field with no default. There is no path in the system that guesses severity from wording, no model that reads a foster note and assigns a level. A missing severity is a form error, not a gap to be filled in. The person who saw the behavior names it.
Review history is immutable
Review history is immutable
A behavior observation's review can be recorded exactly once. A later reviewer cannot silently overwrite what an earlier reviewer decided; the history stands. If marking the observation as affecting the public profile, the reviewer must write a note explaining why, in their own words.
Observation and request stay separated
Observation and request stay separated
Foster check-ins carry an urgency flag so a question does not wait behind routine updates, but the observation and the request are kept distinct. "She startles at the doorbell" is an observation. "Can someone call the vet about this" is a request. Conflating the two is how behavior notes become misleading.
What we do not know yet.
The system organizes context; it cannot tell a team what a finding means.
A dated, attributed medical row is genuinely useful, and it is also genuinely incomplete without a veterinarian reading it. We do not pretend the record is a clinical opinion. Veterinary authority stays with veterinary professionals, full stop.
How much medical context should be public is still an open question.
Adopters deserve honest context; animals deserve not to be reduced to their diagnoses. The profile review gate exists to hold that tension with a person, not resolve it with a rule. We are still learning where the right line sits, case by case, with the vets and behavior professionals who live it.
We have no outcome data.
No shelter has completed a pilot, so we cannot say the medical tooling changes outcomes. What we can say is what it does: keep the context attached, route the follow-ups, and refuse to publish past an unresolved medical question. The rest is a hypothesis we want to test with real teams.
If you are a veterinarian or vet tech and something on this page is wrong, or missing, we want to hear it. The record is shaped by the people who do the work; tell us what it gets wrong about yours.
Bring the medical workflow that keeps breaking.
Intake exams, vaccination follow-up, medication instructions, behavior review: bring one, and we will map it against what the record holds today.