Doris.AI: AI-powered fashion app for personalized style and shopping

Doris.AI, a Brazilian artificial intelligence technology platform that entered the business-to-business (B2B) market in 2023, is now aimed at the end consumer. It is available as an app on both iOS and Android.

Doris.AI has developed an artificial intelligence that uses the clothes a person already owns to help them discover combinations, understand their style and make more contextualised purchasing decisions. After downloading the application, the user simply sends a photo of themselves and interacts with the chat. The app saves existing or purchased items, which form a digital wardrobe over time. When an item is missing, or when a new piece suits the style and occasion, Doris searches for alternatives from different brands and price ranges. It then shows a simulation of the outfit on the user's body, even before the purchase is made.

“Previously B2B, it is now a horizontal AI solution,” says Marcos de Moraes, CEO and the main figure behind this innovation, who gave FashionUnited an exclusive interview. Read it below.

How does the new Doris work?

Anyone can download it for free on both Android and iOS. The user takes photos of the items in their wardrobe, which are then stored. If, for example, you have a special occasion like a wedding party, you click on the app and ask Doris: “What can I wear?” Doris will search your wardrobe and can combine items to show what is possible to wear without buying new clothes or accessories. If the person wants something new, the app captures items from various brands and makes suggestions. This encourages the user to think and see how these pieces look on their body before deciding to invest.

Does Doris partner with fashion brands?

We are testing partnerships with some brands, but Doris has a separate space for each brand. The app combines the pieces the user already owns with new pieces from the brand. It acts as an agent that accompanies the user to find what they need or want.

Doris.AI has a universal browser that can search collections from various brands. To do this, the app goes to the brand's website, reads what is wearable and presents it to the user in the context of their personal taste.

Another service available is asking Doris, for example, which shops in a particular shopping centre stock a certain product. It will provide an answer. In the shop, if the user takes a photo of any item, Doris will immediately combine it with what the user has in their wardrobe, creating all possible outfit combinations.

How much was invested in development and how many people and companies worked to create this ecosystem?

An investment of 40 million reais was made. We have approximately 40 people involved and very strong partnerships, particularly with Oracle and Nvidia, both of which are startups. Although they are large, they think like small companies. They give us a lot of attention.

Is it a product only for Brazil?

No. It is a global product, and we are exploring markets such as the Middle East, among others. In some Middle Eastern countries, for example, where there were no fitting rooms in shops due to religious prohibitions, customers had to take a suitcase of clothes home to try on. So, instead of taking suitcases of clothes from a shop, women in these countries can use Doris to try everything on virtually at home.

How is the app monetised, since it is a free download?

Initially, we earn from the collection of anonymous statistical data, which we use internally. We also earn from purchases (without being pushy, just by suggesting), receiving a commission from the shop. So far, we have noticed very few returns, as the user can virtually try on all the items with minimal error, leading to more confident purchases.

We are now planning a social network where users can post their outfits, and anyone can try on the same outfit. We will then have a paid curation service with subscriptions. This will allow any user to become an influencer (provided their outfits and style are successful), earning from it and sharing the revenue with the app. We plan for this to be launched by the end of 2026.