I'm Keeda — a solo developer, and the person behind Looksy. This page is the honest version of why this project exists, what it's trying to do, and where it stands. No launch-day mythology, no invented numbers. Just the reasoning.
The “where's that from?” moment
Everyone has had the moment: you see someone wearing something great — a jacket, a pair of jeans, sunglasses — and the only question in your head is “where's that from?” You might ask, you might screenshot it, you might scroll past and forget.
The screenshot folder is where clothes go to be forgotten. It's a dead end: a picture with no name, no link, no next step. Looksy is the next step. Instead of a folder of unanswered questions, you get a conversation that ends at a real listing you can actually buy from.
Why Instagram
Because that's where the outfits are. Outfits are shared on Instagram more than anywhere else — in posts, stories, reels, and the DMs where people actually talk about them. If the question is “where's that from?”, the best place to answer it is in the feed where you saw it, in the same app, without leaving for a browser tab and a search bar.
Starting in DMs also keeps it human. You're already messaging; asking a bot to identify a jacket is one more message, not a new app to learn. And the conversation can go both ways — you can clarify, correct, and steer in a way that an empty search box can't.
Why clothes first
Looksy could chase a hundred product categories. It doesn't — not yet, anyway. Clothes first, for three reasons:
- Clothes are visual. A photo carries most of the information needed to search: shape, colour, material, cut. That's a much better fit for image-based identification than, say, electronics, where the useful details are printed in tiny text on the underside.
- Clothes are everywhere. People encounter outfits constantly, in person and on screen, so the “where's that from?” moment happens often enough to matter.
- Clothes are forgiving. When an exact item is gone, a similar style is often a genuinely good outcome — which gives the search room to be honest about uncertainty instead of forcing a false certainty.
What a useful answer should look like
Over time, I've landed on three rules for what Looksy should return:
- Relevant garments, not a wall of guesses. You asked about the jacket. The answer starts with the jacket — not “everything we could see in the photo”.
- Real links, not vibes. Every card leads to an actual listing at a real store. A bot that describes your dream jacket but can't take you to one is a magic trick, not a tool.
- Honest uncertainty. An exact match and a similar style are different things, and the card says which one you're looking at. If I'm not sure it's the exact item, you'll hear “here are similar styles” — never a fake confidence percentage.
A solo project, shaped by feedback
Looksy is built by one person, in the open, and it shows — some edges are rougher than a big team would leave them, and some features are still on the bench. What it has going for it is that every rough edge is fixable, and the fixes come from the people actually using it.
Made by Keeda, a solo dev — feedback welcome. Found a wrong match, a confusing card, or a feature you wish existed? Tell me directly: keedascully@gmail.com. The honest misses are as useful as the hits — probably more.