AlphaCue AI Vet — AI Pet Health Assistant & Vet Consultation Platform
Ask about your pet's symptoms, get an answer in seconds — and reach a real vet when it actually matters.
When a dog stops eating or a cat starts limping at 11pm, the owner's only options are guessing, a panic search that returns contradictory forum posts, or an emergency clinic visit that may turn out to be unnecessary. Meanwhile their pet's actual history — vaccinations, past illnesses, medications — lives in a folder of paper receipts, so nobody, including the vet, has the context that would make the answer accurate. And the obvious technical fix, an AI assistant, has a cost problem: every free user asking the same question burns paid API calls on an answer that was already generated last week.
A Django platform that pairs an AI triage assistant with a real vet-consultation channel. Symptoms are answered by Gemini with the pet's own profile as context — species, breed, age, weight, diet — returning possible causes, an urgency level, and recommendations. A keyword safety layer intercepts seizure, bleeding, unconscious, and breathing-difficulty reports before any AI call or cache lookup and overrides the response with an instruction to seek immediate veterinary care. Cost is controlled by a MySQL-based response cache: inputs are normalised and text-matched against previous queries, free users are served cached answers within a 5-per-day quota, and premium users' fresh API calls are cached to serve future free users. When AI isn't enough, the owner pays for a real consultation and gets a two-way thread with an assigned vet, ending in a structured digital prescription. Built as a lean Django monolith with a strict no-Redis, no-Celery, no-worker constraint so it runs on shared cPanel hosting.
Outcome
Shipped through v0.1 and extended well past the original ten-milestone plan — 13 Django apps, 19 models, and 101 routed endpoints covering web and API. A full JWT REST API was built alongside the session-auth web app (deliberately separated paths) and documented in a 22-page Flutter integration guide, so the mobile client could be built against a stable, mirrored contract rather than reverse-engineered. Went through mobile store review preparation, adding user blocking and legal pages to satisfy App Store guideline 1.2. Payments run through bKash and Stripe; push notifications through Firebase.
Key features
- AI symptom checker grounded in each pet's own profile — species, breed, age, weight, and diet feed the prompt
- Emergency keyword override that fires before any AI call or cache lookup
- MySQL-based AI response cache with input normalisation — free users hit cache, premium calls refill it
- Rolling 24-hour AI quota for free users with a non-blocking upgrade prompt instead of a hard block
- Paid vet consultations — assign a vet, threaded messaging with image attachments, resolution tracking
- Structured digital prescriptions with medicine, dosage, frequency, and duration line items
- Pet profiles with photo upload, content-validated file types, and enforced size limits
- Health record timeline — vaccinations, illnesses, medications, vet visits — with attachments and PDF export
- Structured community Q&A: answers require the experience, the action taken, and the outcome, with upvoting and reporting
- User blocking and moderation tooling built for app-store compliance
- Free and premium subscription tiers with Stripe billing and automatic expiry fallback
- bKash payment flow for consultations in the Bangladesh market
- Firebase push notifications with per-device token registration
- SEO-ready blog with slugs, meta fields, tags, and view tracking
- Dedicated vet panel for consultation queue, replies, and payout tracking
- reCAPTCHA v3 on all web forms; JWT-protected API with rotating refresh tokens
- Mandatory disclaimer on every AI response — informational only, not a replacement for a licensed vet
Tech stack
Build something similar?
Tell us about your project. We'll share how we'd approach it.
Start a project