Aidrobe

Two-person team

Find clothes from different stores and try them on virtually before you buy.

We also have to handle the try-ons that go wrong

If someone is trying on trousers but their photo only shows their face, generating an image isn’t going to help. We check the photo first so they don’t spend credits on a try-on that can’t work.

We also handle the less exciting parts: a result taking longer than expected, a request failing, or credits needing to be returned. Those things matter when you’re the person waiting.

We’re two friends building and running Aidrobe together, from the iOS app to the backend.

Visit Aidrobe

How the product changed

Now
We’re working on how long try-ons take and what they cost to run. We want people to be able to try clothes without a long wait or an expensive AI bill making the app impractical.
Production work
We added photo checks, retries and credit refunds for failed requests. Someone shouldn’t lose their credits because a generation failed.
Discovery
We made finding and saving clothes a bigger part of the app, so you can browse for something you like before trying it on.
Try-on first
We started with uploading a photo and seeing how a chosen item of clothing would look on you.

Handling photos, credits and failures

We run the image-generation work on Google Cloud Run and keep credit balances and subscriptions in PostgreSQL. Keeping the balance on the server lets the app check what someone has available before starting another try-on.

We use Adapty to manage subscriptions and change the offers people see. Pricing needs to cover the cost of generating images, so we work on that alongside the try-on experience.