About

About Pearfit

What Pearfit is

Pearfit tells you what size to buy. It compares the shirt you are thinking of buying with shirts you already own, and tells you which one it fits like. For example: “Get the M. It fits like your olive heavyweight tee.”

Every answer also says how sure the pick is, so you know how much to trust a size before you order.

It is for shoppers who are looking at a shirt online and are not sure which size to get. It works from your own wardrobe, so no body scan is needed, and it is free to try. For now it covers men’s tees.

Brands and stores can also use Pearfit. A pilot starts with fit numbers for one product category, and a size tool on a store’s own product pages is a later step.

What we believe

A size answer should be about fit, not a picture. The useful question is “what size, will it actually fit”, so Pearfit gives a sentence and a confidence level instead of a render.

A size answer should be something you can check. “About 2 cm bigger in the chest than your shirt” is the kind of claim you can verify. “The AI says M” is not.

Your own clothes make a good starting point. A shirt you already know fits you is a reference you know, and it is not tied to one store.

Most size tools only work on big stores that pay for them. We want a tool a shopper can use on any store.

How we keep it honest

Code makes every decision. Language models only turn messy text into numbers. No AI model picks your size. The size comes from comparing measurements with fixed rules.

A missing measurement stays blank. We never guess one, because a made-up number would spoil the whole answer, while a missing one only lowers confidence.

We show how sure we are. Confidence goes down when the data is weaker or when two sizes score almost the same.

Reviews never override the measurements. What reviewers say about the fit can only break a tie between sizes that are close.

We say what the tool cannot know. It matches measurements, so it cannot tell whether you want a particular shirt baggy or slim.

We keep the fit numbers from reviews, not the review text. Full review text is not kept in our shared database, only the fit numbers for each size.

Your wardrobe stays on your device. In the shopper app it is saved in your own browser, and the measurements you type are sent to our size service only to work out the answer, which does not store them.

Pearfit is not affiliated with, endorsed by or sponsored by the brands and stores we name. Their names belong to their owners, and we use them only to say which product a guide or list is about.

A suggested size is an aid for your own decision, not a guarantee. Fit differs between people, products and production runs, so please check the store’s own size information and return policy before you buy.

Who is behind Pearfit

Pearfit is built by four co-founders:

  • Cole Roguski, co-founder
  • Steven Liu, co-founder
  • Jinyu Chen, co-founder
  • Alvin Sheng, co-founder

Contact

Questions, requests, or a brand that wants to talk about a pilot: write to customerservice@pearfit.app.

Tell us what worked, what was confusing, and what you wish Pearfit did, using the feedback form on the home page.