Connect your catalog
Your product list is read at regular intervals. New products are added, sold-out ones removed, price and stock tracked.
AI that understands your ecommerce catalog.
Stilora learns your products from their images and descriptions, then powers semantic site search, similar-product recommendations, complete-the-look outfits and SEO product content.
Products are clustering by meaning…
One intelligence. In every moment of shopping.
Stilora learns a product once. Search, content, similar products, outfit completion and every other feature share the same understanding.
Your product listImage, name, description, price, stock
Order data optionalFor frequently bought together
Stilora core
Reads the productImage and text are analysed together; 20 attributes are extracted from every product: color family, pattern, material, silhouette, formality, season, usage context.
Places it in meaning spaceImage and description merge into a single “meaning”. What the product is, what completes it and how it looks are each known separately; works in 100+ languages.
Applies industry rulesSeparate matching logic for fashion, cosmetics, electronics, home & living, cleaning and general retail.
The product is read
Image and description are analysed together. Details missing from the text are understood from the image.
It enters meaning space
Even attributes nobody wrote down, like “black buckle belt”, become searchable and matchable.
It answers instantly
Products are prepared in advance. Shoppers never wait; results appear right away.
- Attributes read from each product
- 20
- Perspectives per product
- 3
- Supported languages
- 100+
- Industry verticals
- 6
Semantic site search: by intent, not by keyword.
A shopper who types “light-colored dress for a countryside wedding” shouldn’t see an empty page just because no product name contains “wedding”. Stilora matches the meaning of the query with the meaning of your products.
- It parses the query itself. Typing “sandal” searches only shoes; typing “men’s shirt” searches only men’s products.
- It finds details in the image. A thin strap or a diamond pattern is searchable even if the text never mentions it.
- You keep a say in ranking. You can boost campaign products; the effect is limited, so irrelevant products never jump ahead.
Keyword search
Stilora
Illustrative demo. Products and numbers are examples.
AI product descriptions for SEO: unique content for every product. For thousands in a single request.
In most catalogs descriptions are either empty or copied. Stilora reads each product through its image and attributes and writes copy that is ready for search engines and AI assistants.
Raw product in your catalog
Description field is empty.
Content generated by Stilora
- Description
- Features
- Keywords
- Outfit suggestion
Illustrative demo. Real content is generated from your own product data.
For search engines and assistants. Title, description and features are organised for search engines and AI assistants.
Complete fields. Title, description, features, keywords and a short summary shown in search results.
Works at scale. Thousands of products are processed in the background and each piece of content is delivered to your system when ready. Small batches are answered instantly.
Illustrative demo. Scores show how the formula works.
Similar-product recommendations: if it’s not there, show a stand-in.
When a size sells out or a color isn’t right, your customer sees alternatives that feel the same. Substitutes, not complements: boot for boot, lipstick for lipstick.
- Meaning and appearance together. The default hybrid mode weighs both; you can pick just one if you prefer.
- It filters the noise itself. Other sizes of the same product, items no longer on sale and products from other industries never make the list.
- Fast and free of per-call cost. Recommendations are prepared in advance, so they appear instantly and add no extra cost per view.
Complete the look: a jacket, a question and an abandoned cart.
Outfit completion presents a product together with the pieces that complete it. Let’s see why it matters through Elif’s evening.
selvi0
What will I wear this with?
People who viewed this also looked at
Tab closed
Goes well with this jacket
94%
91%
89%
Add outfit to cart · ₺5,146
Is there a more casual alternative?
White leather sneakers also work instead of loafers. Size 38 is in stock.
Outfit added to cart, 4 items
-
10:40 PM
Elif likes a blazer jacket.
She has been browsing the product page for minutes. The jacket is lovely but one question is on her mind: “What will I wear this with?”
-
10:43 PM
A customer with no answer closes the tab.
A classic recommendation box shows her three more jackets. But Elif doesn’t need another jacket; she needs trousers, shoes and a bag. The cart stays empty.
-
With Stilora
The scene changes on the same page.
Trousers, a pair of loafers and a bag matching the jacket’s color, cut and season appear. Why each piece was chosen is explained in a single sentence.
-
Result
One product becomes a look, a look becomes a full cart.
If Elif hesitates she asks the chat stylist, sees which sizes are in stock and adds the whole outfit to her cart with one tap.
Pick a product. Let Stilora build the outfit.
Change the product and the theme of the day; see how the outfit is rebuilt and why each candidate was rejected.
Illustrative demo. Real recommendations are generated from the products and stock levels in your own catalog.
Every outfit passes five rules before it reaches a customer.
- Anatomical completeness
- Someone viewing a top is offered bottoms or shoes; the same body area is never filled twice.
- Color and tone harmony
- Harmonious color families are matched instead of random colors.
- Forbidden pairings
- A bikini never meets trousers and a winter boot never meets summer shorts in the same outfit.
- Season fit
- Four-season pieces are told apart from those that belong to a single season.
- Silhouette balance
- A slim-cut top is balanced with a loose-cut bottom.
Let your customer describe the need, let the stylist listen.
The chat stylist understands budget, occasion and taste. Three agents prepare the answer together.
- Stylist
- Understands the need and builds the outfit.
- Stock
- Checks size and quantity in real time.
- Cart
- Adds the selected products to the cart with a single command.
Chat stylist
More features from the same core.
Frequently bought together
Associations learned from your real order data. When data is thin, gaps are filled with semantic similarity.
Image search
A single photo is enough: it finds the closest match in your catalog and builds an outfit around it.
Single complement
The best single match instead of a full outfit: trousers for a jacket, shoes for a dress.
Quality feedback
Recommendations are scored with your team’s feedback. Weak matches become visible and get fixed.
Every industry has its own outfit.
A lip liner that matches a lipstick, a case that fits a phone, a cushion that suits a sofa. Each vertical works with its own attributes and matching logic.
Two use cases
A cosmetics brand
Cosmetics and personal careIn cosmetics, “similar” doesn’t just mean color; skin type, skin concern, formula, ingredients, tone family and undertone are weighed together. Here, outfit completion becomes a care routine.
- Cleanser
- Serum
- Moisturiser
- Sun care
- Lips
- Search
- “night cream for dry skin”
- Similar products
- lipstick alternatives in the same tone family
- Outfit completion
- a routine from cleanser to sun care
- Content
- ingredient and usage-focused descriptions
A brand that only sells shoes
Single-category catalogEven with no tops or bottoms in the catalog, Stilora keeps its value. Similar products, search and content stand out here.
- Selected
- 0.90
- 0.88
- 0.81
- Low
- Similar products
- the closest alternative to a sold-out model
- Search
- “boots to wear on a rainy day”
- Content
- unique descriptions for thousands of models
- Outfit completion
- with the bags and care products in the catalog
Live in a few steps. Secure at enterprise scale.
You don’t need to rebuild your site. Share your catalog; Stilora learns it and shows results on your site, in your own design.
Share your catalog
Connect your product list. New products, price and stock changes are followed automatically.
Stilora learns your products
Every product is read through its image and description; details like color, material, season and usage are extracted.
Shoppers see it on your site
Your search box, product pages and chat area get smarter. The look stays entirely in your own design.
Every brand in its own space
Each brand’s data is kept separate. Brands never see each other’s data.
Secure access
Only authorised people and requests from your own site are accepted; every action is logged.
Instant answers
Products are prepared in advance, so your shoppers never wait.
Everything on one screen
Monitor performance, errors and cost from a single management dashboard.
Frequently asked questions
For any other question, get in touch with us.
What is Stilora?
Stilora is an AI platform built by Inveon for ecommerce brands. It learns your product catalog from images and text, then uses that understanding for semantic site search, similar-product recommendations, complete-the-look outfit recommendations, a chat stylist and SEO product content generation. It works with fashion, cosmetics, electronics, home & living and general retail catalogs.
Do we have to use all of Stilora’s features at once?
No. Stilora is a single product; all features use the same product data and the same core. You decide where on your site each feature appears. For example, you can start with search and content and add outfit completion to your product pages later.
Does it recommend products outside our catalog?
No. Stilora only searches and recommends products from your own product list. Every result links to a purchasable product page.
Can an out-of-stock product be recommended?
Your product list is read at regular intervals and products no longer on sale are filtered out of results. The chat stylist checks size and quantity in real time.
Our product descriptions are incomplete; will it still work?
Yes. Stilora reads the product together with its image; details missing from the text are understood from the image. You can also fill in missing descriptions with SEO content generation.
We’re not in fashion. Is it right for us?
Separate matching logic is ready for cosmetics, electronics, home & living, cleaning & care and general retail. Even in single-category catalogs, like a brand that only sells shoes, similar products, search and content keep their full value.
We have more than one brand. Can we manage them all?
Yes. The infrastructure was designed multi-brand from the start. Each brand works with its own data space and its own settings; all are monitored from a single management dashboard.
Which ecommerce platform does it work with?
It is platform-independent. You share your product feed; results are added to your site through a ready-made widget or the API. No rebuild is needed and the look stays in your own design.
Introduce your catalog to Stilora.
Connect your product list and see how your catalog is understood, from search results to product pages.