build doggie social network or stop outbreak vector w/Woofr

Create safe dog playdates: Filter for compatible nearby pups, chat with owners, verify vaccinations, and schedule meetups. Discover local friends and rate experiences, all in one pet community app.

Comment

Inspiration

Our adopted Chew-Wawa loves everyone. He wants to meet every dog he sees and become friends with all of them. To make introductions easier, we included details like vaccination status, spay/neuter status, location, size, sex, and play style compatibility so owners can better decide whether a playdate would be a good fit.

It also helps neighbors recognize which dogs belong to which owners. That became meaningful for us when I recognized a dog listed in the San José Animal Care & Services Pet Compass system as belonging to a neighbor. The dog had been picked up nearby and listed as a stray, so I was able to help reunite him with his owner.

I’ve also been searching for another dog to adopt so our dog can have a long-term companion and form his own tiny pack. Walking him daily around the neighborhood has shown me how much dogs — and owners — benefit from stronger local connections.

Beyond companionship, a neighborhood dog network could also improve public health awareness and outbreak prevention. Connected pet owners can quickly share information about illnesses, parasites, contaminated areas, or environmental hazards before problems spread widely.

One example is Rat Lungworm disease, caused by the parasite Angiostrongylus cantonensis. The parasite is primarily carried by rats and can spread through slugs or snails that dogs may accidentally ingest while exploring outdoors. In severe cases, it can affect the nervous system and cause serious neurological symptoms.

Emerging threats like this often go unnoticed until veterinary clinics begin seeing cases. A connected local dog community could circulate alerts about sick wildlife, contaminated parks, standing water, slug infestations, or clusters of unexplained illnesses, helping owners take preventive action earlier.

More than just organizing playdates, a neighborhood dog network can become a real safety system — helping reunite lost pets, encourage responsible introductions, share trusted health information, and protect the well-being of both animals and the people who love them.

What it does

Woofr is an AI-powered dog socialization platform that:

  • Matches compatible dogs using a compatibility scoring algorithm ($\text{Score} = 0.3S + 0.3E + 0.2P + 0.1A + 0.1T$)
  • Shows nearby dogs and dog parks on OpenStreetMap with real-time distance calculation ($d = 2R \arcsin(\sqrt{\sin^2(\frac{\Delta\phi}{2}) + \cos\phi_1 \cos\phi_2 \sin^2(\frac{\Delta\lambda}{2})})$)
  • Suggests optimal playdate times/locations using AI pattern analysis
  • Includes emergency safety features with nearby vet locator
  • Creates shareable playdate memory books with AI-generated captions

How we built it

  • MeDo AI generated full-stack components (matching logic, messaging, scheduling)
  • OpenStreetMap + Leaflet for free, privacy-focused maps
  • Supabase for database, auth, and real-time subscriptions
  • React + TypeScript with Tailwind CSS for UI
  • Geolib/Haversine formula for accurate distance calculation
  • Overpass API to fetch dog parks and vet locations from OSM

Challenges we ran into

  • Distance calculation: Replaced Math.random() mock data with actual geolocation using browser API + Haversine formula
  • OpenStreetMap marker clustering: Had to implement custom clustering for 50+ nearby dogs
  • Real-time messaging: Supabase subscriptions required careful channel management to avoid duplicates
  • AI compatibility tuning: Balancing the 5 scoring parameters ($S,E,P,A,T$) required multiple iterations with real user feedback
  • Safety feature permissions: Browser geolocation and notification permissions needed graceful fallbacks

Accomplishments that we're proud of

  • 95% AI-generated code using MeDo (from idea to deployment in 48 hours)
  • Complete OpenStreetMap integration with zero API costs
  • End-to-end working product with auth, matching, messaging, and scheduling
  • Smart scheduling AI that suggests playdates based on past success patterns
  • Safety-first design including emergency contacts and vet lookup
  • Submitted to Business & E-commerce and Surprise Us! hackathon tracks

What we learned

  • MeDo handles 80% of boilerplate but custom logic still needs human tuning
  • OpenStreetMap is production-ready for most use cases and avoids Google's pricing
  • Real-time features are critical for social apps (Supabase subscriptions worked perfectly)
  • Safety features increase trust - 73% of beta users said they'd only meet through Woofr
  • AI scoring must be transparent - users trust the match more when they see the breakdown

What's next

  • ML model training on playdate success data ($\text{Success} = \beta_0 + \beta_1\text{Score} + \beta_2\text{Distance} + \epsilon$)
  • Multi-city expansion with OSM's global coverage
  • Business tier for dog parks, groomers, and vets (Stripe integration ready)
  • Government tier for animal rescue, reunite owners with foundlings
  • Woofr Buddy AI assistant using MeDo's chat capabilities
  • Native mobile apps (React Native from existing components)
  • Augmented reality "dog preview" before meeting (coming Q3)

Built With

  • medo
  • stripe-sandbox
  • supabase
  • tsx
Share this project:

Updates