Dog-e-dex

Dog-e-dex turns every dog you meet into a collectible. Point your iPhone at a pup, snap a photo, and the app uses an AI vision model to identify the breed and add it to your personal "dex." Over time you build up a catalog of every dog you've encountered — friends' pets, neighborhood regulars, dogs at the park — complete with breed information and the photo you took. The mechanic is borrowed shamelessly and lovingly from Pokedex: see one, photograph it, log it, repeat.

It's a small, joyful idea executed with care — exactly the kind of thing that benefits from being a polished iOS app rather than a website.

How it was built

Dog-e-dex is the work of Cynthia Chen, a product designer at Block who, by her own account, had no engineering background when she started. She built the app over roughly two months by vibe-coding her way through Swift and SwiftUI with a stack of AI assistants. Claude did the heavy lifting as her primary collaborator, with Replit, ChatGPT, and Cursor filling in around the edges for prototyping, debugging, and getting unstuck.

Cynthia's experience became one of the most-cited stories of the vibe-coding moment in 2025. In coverage by DNYUZ and Yahoo Tech, she described the process as one part designer-led product work and one part "babying" the AI through the parts where it kept getting things wrong — a tip that resonated with a lot of non-engineers trying to ship their first app.

Why it matters

Dog-e-dex is a clean proof that a designer with strong product instincts can ship a real, App-Store-quality iOS app without ever having shipped production code before. The app itself is delightful; the meta-story — two months, no traditional engineering background, a Pokedex for dogs — is part of why it kept getting written about.