PawDoc
AI-powered pet health triage app that uses computer vision to assess symptoms and guide emergency care decisions.
"Monitor at home" or "go to the clinic now." The triage framework is the moat — the AI is the surface.
Tech Stack
Mobile
- React Native
- TypeScript
Backend
- Node.js
AI
- Computer Vision
- Multimodal AI
Overview
Why
Pet owners search symptoms at midnight, find Reddit threads of varying quality, and either panic-drive to an ER or wait too long. The willingness to pay for a calm, structured, evidence-grounded triage opinion is high — and the addressable market (170M US pet households, 72% treating pets as family) is real.
How
Multimodal AI reads photos + symptom descriptions and scores them against a structured veterinary triage framework. The output is a prioritised action recommendation — "monitor at home" / "book a routine appointment" / "go to an emergency clinic now" — explained in terms of observable signs, not speculation. Three user archetypes — First-Time Pet Owner (anxious), Busy Professional (decision-driven), Budget-Conscious Owner (cost-aware) — shape the prompt and the response framing.
Trade-offs
Liability is the central trade-off. The triage framework defaults conservative — the cost of a false-negative ("monitor at home" when the pet needs the ER) dwarfs the cost of a false-positive. Real telemedicine integration (Phase 2) requires veterinary licensing infrastructure we don't yet have. The roadmap is intentionally staged: ship the symptom checker first, validate willingness to pay, then layer the regulated surfaces.
