The challenge
We started with what dog owners love: breed recognition of all the dogs you encounter on the street. But we quickly discovered the real value was elsewhere. Dog owners struggle daily with scattered information and time-consuming processes for their own dog - scheduling appointments takes too many steps, contact details are scattered, and simple actions are unnecessarily cumbersome.
Our approach
We built Snuffl as a full-fledged stand-alone app with dual purpose that serves both needs. On one hand, quick breed recognition when you're curious about a dog's breed, on the other hand, a central place for everything about your own dog.
Instead of having users type everything, we opted for workflows where you mostly click. Adding contacts happens via Google Places integration, so your vet or groomer is added in two clicks.
A concrete example: if you want to call the vet, you open the app, see the dashboard, and click the phone number. Two clicks total from the moment you open the app.
Product flow
Scan and analyze
Get an indication of breed/mix.
Create a profile
Smart input assistance, mostly clicking.
Manage contacts
Add contacts in 2 clicks via maps integration.
Manage calendar
Choose appointment type and link your vet directly.
Scan archive
Detailed overview of all scans with analyses.
Export to calendar
All appointments to your personal calendar in one click.
The result
The core functionality was developed in a short time with AI as co-developer, strategically switching between ChatGPT, Claude and Gemini. Thanks to Claude Code, we could implement six languages in just one day.
The app is now live with OAuth integration, Google Places API connection and multilingual content. Users come in for breed recognition and discover the practical value of the hub functionality.
Why this pattern works for organizations
The Snuffl model combines curiosity with practical value. You start with what people enjoy and lead them to what they really need. By removing all friction from the user experience, complex workflows suddenly become simple.
The AI-accelerated development method made ambitious features realistic that would otherwise be unattainable.
ChatGPT, Midjourney, Photoshop Firefly
React + Tailwind CSS
Supabase, Edge Functions, Row Level Security
OpenAI GPT, Google Places API, OAuth
ChatGPT, Claude Code, Gemini
Check out these cases
Interested in a similar approach?
Want to try Snuffl? The app is live at snuffl.app
Want to discuss a similar application? Get in touch.