The integration addresses a fundamental friction point in live commerce: unlike static e-commerce catalogs, Whatnot’s inventory and buyer intent fluctuate rapidly. Emmanuel Fuentes, VP of Data and AI at Whatnot, noted that the platform currently processes over 500,000 hours of live video and millions of interactions weekly. By folding in Shaped’s technology, the company intends to shrink recommendation latency from minutes to near-instantaneous, ensuring that search results remain relevant as auctions begin and end.
Whatnot acquires machine learning startup Shaped to fuel live commerce
To solve the volatility of live-streamed auctions where inventory shifts by the second, Whatnot has acquired Shaped, a machine learning firm specializing in real-time discovery. The deal brings a team of AI researchers to the platform, aiming to transform how millions of buyers navigate an ever-changing digital marketplace.
Shaped founder Tullie Murrell, a former Meta engineer, will now head Whatnot’s new Applied AI Research group, bringing nearly a dozen engineers along to spearhead the initiative. This acquisition follows a period of aggressive expansion for Whatnot, which recently surpassed one billion orders and secured a valuation exceeding $11 billion after a $225 million Series F round. As the company pushes into dozens of new product categories—ranging from fine art to vinyl—the ability to deliver hyper-personalized discovery is becoming a primary competitive advantage against rivals like eBay and Poshmark.
Comments (0)
No comments yet. Be the first!