When Svetlana Kordumova was studying for her PhD in AI and computer vision, she became frustrated with the process of searching for items to buy online. Search results were often inaccurate, and she knew the technology she was learning could improve the experience. Pixyle AI launched in 2019 to improve product discovery on e-commerce sites and today announced a €1 million (approximately $1.05 million USD) seed round from South Central Ventures.
The startup, which has offices in Amsterdam and North Macedonia, now works with more than 20 clients, including Depop, Otrium and Minto. Over the past three years, it has tagged more than 250 million images and says it has increased conversions for its retail customers by an average of 10%.
Pixyle AI’s neural networks train its visual AI algorithms to not only identify fashion items in an image, but also categorize them based on attributes, such as color or pattern, that match the keywords shoppers use when searching for a product. article. The aim is to ‘see’ images as a human would. For example, someone searching for a “short summer dress with floral print in pink and purple” will get results with all those attributes.
Kordumova, who obtained her PhD from the University of Amsterdam, first created a visual search app for consumers before switching to B2B in 2019. She told londonbusinessblog.com that one of the biggest challenges online retailers face is cart abandonment, often due to poor site search. and product discovery. Research from Google Cloud shows that while more people than ever are shopping online because of the pandemic, 52% will abandon their shopping cart and go to another site if they can’t find just one item.
The reason for search results is usually bad data. Retailers often get incomplete and inaccurate product data on brands from people who list used items for sale, which means items don’t show up in search results. Many retailers solve that problem by manually entering better product data, but that process is labor intensive, expensive and prone to human error.
“Take the example of color attributes, what one person might rate as yellow, another person might find more orange,” Kordumova said. “In the case of second-hand marketplaces with millions of products uploaded to the platform, it is simply an impossible task to manually add attributes to the metadata.”
Pixyle AI automates the process of extracting detailed attributes from images and now has a growing fashion taxonomy that already clocks in at over 20,000 attributes, aiming to cover every possible query about clothing.
The startup’s clients include online marketplaces, brick-and-mortar retailers, and fashion tech startups such as clothing catalog app Wherering, virtual fitting solution Virtusize, and live shopping marketplace Galaxy. Pixyle AI has helped brands moving from brick-and-mortar stores to “phygital,” or an omnichannel strategy that combines e-commerce with brick-and-mortar stores, by automating product tagging. This increases the speed at which they can digitize their shopping experience.
Some examples of how Pixyle AI’s technology has been used include automating manual product entry and catalog standardization at Otrium. The end-of-season fashion marketplace used to manually tag and process product attributes, but couldn’t keep up with their growing inventory. Kordumova says Otrium was able to improve its productivity by 65% after implementing Pixyle AI to automate color detection for its inbound logistics team.
For consumers, Pixyle AI offers a visual search that allows them to upload an image of what they are looking for and get similar results. Kordumova says sustainable fashion marketplace Project Cece reported a 50% increase in conversion rate to product outlinks after adding Pixyle AI visual search to its site.
Other companies that have developed visual AI tools for product discovery include Syte, Visenze, Vue.AI, and Google, which recently launched a multi-search tool that allows people to search using text and images at the same time. Kordumova says Pixyle AI sets itself apart by focusing on the enrichment of product data with detailed attributes and providing its customers with a high degree of customization and tagging flexibility.
“In order to really make product data enrichment work for each specific situation of our customers, we first have our teams aligned with what we are trying to achieve, making sure we set up the right configurations before our AI models get to work,” she said. say. “This means mapping taxonomies, configuring cloud architectures, and deploying customer and technical support teams to meet our customers’ exact needs, ensuring successful deployment and use of our platform to help them achieve long-term business goals. to achieve.”
Pixyle AI plans to use its new funding to improve its product offerings, expand in the United States and Europe, and move into new verticals. It adds new suites for industry segments and new offerings such as product description generation and label detection using OCR technology that recognizes brands, material composition and size. It will also add “shop the look” and “multimodal” search to its visual discovery product. For verticals, Pixyle AI plans to move into homewares and furniture by the last quarter of 2023.
In a statement about the investment, Jan Kobler, managing partner of South Central Ventures, said: “A critical part of attracting online shoppers is product search, making it easy and quick to find what you want. However, search is hugely underserved and remains an unmet need for retailers and shoppers to date. Pixyle AI is laser-focused on this opportunity and is already turning the knob with more sales for retailers. They have built a robust tech stack, which has been tried and tested in the market and ready to scale.”