Back to all posts

Paws & Co. Veterinary: an AI receptionist after hours

Dr. Sanjana Verma kept her personal phone on every night for pet emergencies — most of which weren't emergencies. AIVA sorts that out before she has to.

MN
Meera Nair
Customer Success

Paws & Co. Veterinary runs two clinics in Bengaluru — one near Indiranagar, one further out toward Whitefield — four vets, closing at 8pm most nights. Pets don't keep those hours, and after-hours coverage is a genuinely different problem than daytime coverage. Before AIVA, an after-hours call at Paws & Co. went to a voicemail box nobody checked until morning — or to Dr. Sanjana Verma's personal cell, which she kept on every night because the alternative felt irresponsible.

Illustrative example. Paws & Co. Veterinary is a composite scenario built to show how AIVA works for this kind of business — not a verified customer account.

The 2 a.m. problem

The trouble wasn't call volume — Paws & Co. was only getting 8 to 10 after-hours calls a week. It was that there was no way to sort them. By Dr. Verma's own estimate, maybe one in five was a genuine same-night emergency: a toxic ingestion, a dog hit by a car, a cat in visible distress. The rest were real questions from worried owners that could safely wait until morning — but a panicked pet owner on the phone at 11pm has no way of knowing which category they're in, and neither, in thirty seconds over the phone, did anyone else. So every call got treated like it might be the urgent one, including the ones that weren't, and Dr. Verma was the one absorbing that uncertainty every night.

They'd tried a general after-hours answering service for a few months before AIVA, and it solved almost nothing. The operators were polite and reliable about picking up, but they had no veterinary background and no way to actually sort a call — so they escalated nearly everything to Dr. Verma anyway, just a few minutes later than if she'd been called directly. The filtering problem just moved one step downstream instead of getting solved.

One case stuck with her: a 1am call about a dog that had eaten a small piece of chocolate — the kind of case poison control references classify by weight and chocolate type for exactly this reason — genuinely low-risk given the dog's size and the amount, but indistinguishable, over the phone in the dark, from the kind of ingestion that isn't. She drove in anyway, because she couldn't be sure. The following week she fielded three more calls that sounded almost identical. Against that, she also remembered the calls where the instinct to drive in had been exactly right — a dog showing early signs of bloat one night, a condition that can turn fatal within hours, where the time she saved by not waiting for morning genuinely mattered. That contrast was the whole problem in miniature: she couldn't tell the two kinds of calls apart in advance, so she treated every one like it might be the second kind.

How the triage works

AIVA Voice now answers every after-hours call and runs through the same intake questions Paws & Co.'s own vets use to open a case: what's the symptom, how long has it been going on, is there a known ingestion, does the animal's age or breed raise the risk.

We built that question set with Paws & Co.'s own vets over about two weeks before going live — walking through past cases, including a few Dr. Verma remembered getting right only on gut instinct at 1am, to make sure the triage logic matched what an experienced vet would actually flag as urgent. That process caught a real gap before launch: an early version of the logic classified a lethargic senior cat as non-urgent, treating "lethargy" as a mild, wait-until-morning symptom on its own — but Paws & Co.'s vets flagged that lethargy in an older animal carries real weight that it doesn't in a younger one, and a case like that shouldn't wait. We added age and breed-risk weighting to the triage logic specifically because of that catch, before it ever handled a real call.

AIVA doesn't diagnose and doesn't give medical advice — that boundary is fixed, not configurable. What it does is figure out whether a call needs a vet right now. If it does, it connects the caller directly to the on-call number with the intake details already gathered. If it doesn't, it reassures the owner, avoids anything clinical, and books the first available slot the next morning.

In practice, the two kinds of calls look very different now. A caller whose dog got into something toxic gets a fast, focused set of questions — what, how much, when — and is connected straight to the on-call vet, who already has that context before picking up, instead of asking the same panicked owner to explain everything again. A caller worried about a day of mild diarrhea gets reassured in plain, non-clinical language, booked into the first slot the next morning, and doesn't spend the night wondering if anyone actually heard them. Non-urgent bookings get an SMS confirmation on the spot for exactly that reason.

The results

Nine weeks in:

  • After-hours calls answered live: voicemail/inconsistent → 100%, every call, first ring
  • True emergencies correctly escalated to the on-call vet: every case Paws & Co.'s own team retrospectively flagged as urgent had already been escalated by AIVA
  • Dr. Verma's personal after-hours calls: down from roughly 8 a week to about 2 — the ones that actually needed her
  • Non-urgent callers booked into next-morning slots automatically: the majority of after-hours volume

The other three vets at Paws & Co. feel it too, even though it was Dr. Verma's phone doing most of the ringing before. Nobody's quietly dreading a shift's on-call rotation the way they used to. It's also the same pay-as-you-go pricing around the clock — Paws & Co. isn't paying extra for the 11pm-to-8am stretch that used to be the hardest hours to staff.

I didn't know how much of my sleep I'd traded for a filtering problem until something else was doing the filtering.

What Dr. Verma tells other vets thinking about this

Build the triage logic with your own clinical team, not a generic template — the lethargic-cat catch only happened because Paws & Co.'s vets were the ones reviewing the questions, not an outside default. Test it against your hardest real cases before it goes live, especially the ones you remember getting right only on instinct. And keep the no-diagnosis boundary hard, on purpose — the moment a system starts guessing at what's wrong instead of sorting urgency, it's doing something different and riskier.

If your practice runs a similar after-hours setup, our guide for veterinary clinics covers the same triage approach, and our piece on how escalation actually works goes deeper into where AIVA draws that line and why it doesn't move. You can also just try it against a few of your own toughest after-hours calls and see where it lands.

Dr. Verma still keeps her phone nearby at night. She just isn't the one deciding, at 1am, whether a call is worth waking up for anymore — the sorting happens before it ever reaches her. The other three vets rotate on-call duty now too, something that felt riskier to hand off before the triage logic existed and nobody was eager to volunteer for. Paws & Co.'s two weeks of case review before launch is roughly the investment other practices should expect — less than that, and gaps tend to surface on a real call instead of during testing, the way the lethargic-cat case almost did.

Share
MN
Written by
Meera Nair
Customer Success

FAQ

Common questions.

By asking the same intake questions an experienced vet would — symptom, duration, known ingestion, and risk factors like age or breed. Paws & Co. built that question set with its own vets, testing it against real past cases before it ever took a live call.

No — that boundary is fixed, not configurable. It figures out whether a call needs a vet right now; it doesn't guess at what's actually wrong.

It connects the caller straight to the on-call vet, with the intake details already gathered — so the vet isn't asking a panicked owner to explain everything a second time.

Volume wasn't Paws & Co.'s problem — sorting was. Even a handful of after-hours calls a week is enough to disrupt sleep every night if there's no way to tell which ones actually need a vet immediately.

Paws & Co. spent about two weeks working through real past cases with their own clinical team before going live — including a gap in how age and breed risk were weighted that only surfaced by testing against a hard real case.

It scales down fine — the triage logic matters more than clinic size. A solo practitioner fielding after-hours calls has exactly the same filtering problem Dr. Verma had, just at a smaller volume.

They get reassured in plain, non-clinical language and booked into the first available slot the next morning, with an SMS confirmation on the spot — nobody's left wondering if anyone heard them.

Like this? Get more.

One email a month. Engineering deep-dives, product launches, customer stories. No fluff.

4,200+ subscribers. Unsubscribe anytime.