# Styel

> Styel is an AI stylist a fashion store runs on its own storefront: a chat widget that reads a shopper’s request in plain language, searches that store’s live catalogue, and answers with three picks and a reason for each.

Styel turns your catalogue into a stylist — three picks, a reason each, on your own site. Fashion shopping should feel like being styled, not searched.

## Key facts

- Styel is a chat widget a fashion store embeds on its own storefront, not a marketplace or a separate destination site.
- Shoppers describe what they want in ordinary language — "black blazer, under $300, structured" — instead of using filters.
- Every answer is at most three products, never a wall of results, each with a short note on why it works and a link to buy; asking for a whole outfit instead returns one piece per part of the look.
- Results only ever come from the store that installed Styel, so a shopper never leaves for a competitor.
- Styel is in early access; the marketing site collects a waitlist and runs a live demo of the real widget.
- Styel is a stylist rather than a search engine: search retrieves and ranks, while Styel decides and explains. It does not replace a store’s search bar.
- Three products get called an "AI shopping assistant": smarter site search, support chat with a shopping bolt-on, and curation apps a shopper installs. Styel is the curation experience, running on the merchant’s own storefront rather than across every store at once.
- Styel is grounded in the store’s live catalogue — sizes, stock, fabric, cut — so it does not recommend something the store has sold out of.
- Styel expands a request into fit, occasion, silhouette, fabric and dress code before it retrieves anything; that apparel model is built for fashion specifically rather than adapted from a general one.
- Every conversation records what was shown, what was clicked, at what rank, and how long the shopper took to decide. Styel publishes no conversion-lift percentage until a holdout test produces one.
- Styel puts no third-party tracker on the storefront and budgets page weight on every release.
- The name is spelled Styel — a deliberate respelling of "style", not a typo. The site calls the assistant a "styelist" for the same reason. The site is styel.ai.

## Site sections

The site is a single page; each section below is a fragment of https://styel.ai/.

- [Meet your Styelist](https://styel.ai/#hero): What Styel is, in one line, plus the waitlist and the live demo.
- [What is Styel](https://styel.ai/#what): The plain-language explanation: what Styel does for a shopper, in the page’s own words.
- [How it works](https://styel.ai/#how): The four steps, from the corner of the page to three finished picks.
- [See it live](https://styel.ai/#demo): A working demo of the exact widget a store’s shoppers see, running on real inventory.
- [Doesn’t this already exist?](https://styel.ai/#category): The three products that share the phrase “AI shopping assistant”, and which one Styel is.
- [Built for fashion, not for everything](https://styel.ai/#specialist): Four claims that separate a fashion specialist from a horizontal assistant.
- [Our rules](https://styel.ai/#philosophy): The three product rules: three picks only, think before answering, the shopper decides.
- [Questions, answered](https://styel.ai/#faq): What Styel is, where its results come from, why a search returns three, and how to get it.
- [Early access](https://styel.ai/#waitlist): Join the waitlist — spots are limited while Styel is in early access.

## Questions and answers

As stated on the page itself, at https://styel.ai/#faq.

### What is Styel?

Styel is an AI personal shopper for fashion stores — a chat widget a store adds to its own site. A shopper says what they are after in plain language, and Styel searches that store's racks and comes back with three picks.

### How is it different from search and filters?

Filters only work if you already know the words the catalogue uses. Styel reads intent instead: something breathable for the gym, dinner but quiet luxury, a blazer that is not boxy. It works out what that means and searches on meaning rather than on keywords.

### Doesn't Google already do that?

Google is built to return everything, ranked. That is the right design for finding a known item and the wrong one for a wedding in Lisbon in September, nothing strapless. That shopper does not want more options, they want the decision made. Styel is grounded in your live catalogue instead — sizes, stock, fabric, cut — so nothing comes back that you sold out of on Tuesday, and nothing comes back that is not yours. And Google is not on your site; neither is ChatGPT. They hand the shopper to your storefront to finish, so whatever meets them when they arrive is what decides whether the visit converts.

### We already have search. Does Styel replace it?

No, and it should not. Search matches words to products. It has no idea what a column silhouette does to a body, that bouclé is a winter fabric, or that wedding guest is an occasion, a hemline and a dress code at the same time. That knowledge is an apparel model, and it does not arrive with a search bar — it is a separate layer entirely. We build it in, for fashion specifically, and put a stylist in front of it.

### Our platform includes an assistant for free. Why pay for another?

Because they are for different jobs, and every store should have the free one: search the catalogue, answer a policy question, complete the purchase. What it does not have is an apparel model, an editorial constraint, or any measurement — it cannot tell you whether it sold anything. Free tells you what something costs. It does not tell you what it returns.

### Isn't this just a chatbot?

A chatbot answers questions. Styel has to make a call. Before the catalogue is touched, what the shopper said is broken into what a stylist would hear in it — occasion, silhouette, fabric, cut, dress code, sizing, and the negatives too, so nothing strapless pushes strapless pieces down the ranking instead of matching them on the word. What comes back is three picks, each with a reason. That cap is enforced in the code that assembles the reply rather than left to an instruction a model can talk itself out of, and a card only renders for a product that genuinely exists in that store's catalogue — so nothing is padded and nothing is invented. A whole outfit is the one exception, at one piece per slot. Then every row is measured: what was shown, what was clicked, at what rank. Restraint is the product.

### Where do the products come from?

Only from the store that installed Styel. Every search is scoped to that store's own inventory, so a shopper is never handed a competitor's product and never sent somewhere else to buy it.

### Why only three picks?

Because fifty results is the problem, not the answer. Three is few enough to actually decide between, and each one arrives with a short note on why it works and a link to buy it. Asking for a whole outfit is the exception: there Styel returns one piece per part of the look, so a head-to-toe answer can run to four or five.

### Can I see it working first?

Yes. The Styel on this page is the real widget running on a real product catalogue, not a video or a mock-up. Open it in the corner of the page and ask it anything.

### How do I get Styel on my store?

Styel is in early access with a limited number of spots. Leave your email on the waitlist and we will get in touch when one opens.

## Contact

- [Email Styel](mailto:hello@styel.ai): partnerships, early access, and press.

## Optional

- [Full site text](https://styel.ai/llms-full.txt): every section of the page as plain text, for when the summary above is not enough.