Almost a year after raising a $50 million megase round, e-commerce veteran Julie Bornstein startup Daydream is releasing an AI-powered chatbot for fashion-focused shopping.
After testing the product with selected users, the company is releasing a chatbot to all users in the public beta today. People can sign up for chatbots. In the chatbot, you will ask if you have a name, date of birth, price range to shop for, or brand preferences.
You can enter queries such as “I want you to wear it at my wedding in the summer in Paris,” upload images, add context via text to search for clothing.
Users can save or improve searches any items in the collection they have created by typing into the chatbot on the left. If you like items but want to change some aspects such as color and style, you can tap the “more” button that appears in one of the items to edit the search.
Feedback obtained from the app will store a variety of items based on the parameters provided by users during onboarding. Daydream creates a Style Passport that promotes many proposals. The web app also shows daily inspiration for items and accessories that suit your taste.
Currently, Daydream does not have an integrated checkout flow, so when a user clicks on an item, they are redirected to the merchant's website to complete the transaction. Startups are reducing the percentage of each sales. Daydream said there were over 8,000 brands on the platform at the time of launch, and the company is onboarding new merchants for free.
Bornstein, who holds executive positions for companies like Nordstrom, Urban Outfitters, Sephora and Stitch Fix, said over the past year the company has brought catalogs of various brands to one location and is working on technology to improve searches to suit AI.
Image credits: Daydream
“I worked in e-commerce throughout my entire career, so searching has always been a forgotten kid and it didn't work very well. In a way, people were trained to be very narrow in the way they search for something in the fashion world.
“But once GPT is launched, we've started to train consumers in how to think about the possibilities of prompts, and what we're trying to do is help you ask what you want to ask.
Maria Belusova, who joined the company as CTO this year, said Daydream does a lot of work to understand the nuances of the items in the catalog. She said traditional searches only showed shopper items based on tags that match keywords that don't work in today's world where customers want longer queries.
“We do quite a bit to understand the details of the product, such as the people who wear this dress, the social attributes such as guests at the wedding, etc., such as the people who wear this dress, etc., the guests at the wedding.
Image credit: DaydreamImage credit: Daydream
The following year, Daydream will allow users to provide clearer feedback with tools such as “Don't show your 4-inch heel.” We will also experiment with a feature that allows users to request a good match with existing items, including personalized suggestions. Additionally, users want to lean towards the social sharing aspect by sharing saved items with friends and family and suggesting they buy them. Another feature that Daydream thinks about is to retrieve an existing collection of another user and use AI to modify it to suit your own needs.
The Daydream team has a long-standing e-commerce experience and focuses on fashion, but startups like Deft and Cherry are also building multimodal searches for shopping. Meanwhile, tech giants like Amazon and Google are focusing on the ability to use AI to search multiple sites and find the right items for their users.