Log in
Skip to content
Back

Notes & guides

Whatnot acquires Shaped to power real-time live shopping recommendations

Whatnot acquires Shaped to power real-time live shopping recommendations

Whatnot Acquires Shaped: A Game-Changer for Live Shopping

In a move set to redefine the live shopping experience, Whatnot, the leading livestream shopping platform, has acquired Shaped, a machine learning startup specializing in real-time recommendation and search technology. The acquisition, announced on July 15, 2026, brings Shaped's founder and CEO, Tullie Murrell, along with nearly a dozen engineers and researchers, to Whatnot. This strategic move aims to supercharge Whatnot's AI capabilities, particularly in discovery and personalization, ensuring that shoppers find exactly what they want—faster than ever.

Why Real-Time Recommendations Matter in Live Commerce

Live shopping is a different beast from traditional e-commerce. Inventory changes by the second, shows can end abruptly or run for hours, and buyers are clamoring for rare items before anyone else. In this high-speed environment, recommendations must be not only accurate but also lightning-fast. As Tom Verrilli, Chief Product Officer at Whatnot, explains, "The latency on its recommendations is now down to just minutes," allowing Whatnot to quickly ingest information—like a product selling out—and adjust recommendations in real time. This speed is crucial because, as Verrilli notes, "Inventory changes second by second. We actually don't know when a show is going to end."

Shaped's Technology: The Engine Behind Real-Time Personalization

Founded in 2021 by former Meta engineer Tullie Murrell, Shaped has built a reputation for its real-time retrieval engine for search, feeds, and agents. The company's technology has been used by hundreds of marketplaces and content brands, including Vox Media and QVC, to power personalized experiences. For instance, Shaped helped New York Magazine develop a more personalized news feed. Now, that same technology will be integrated into Whatnot's platform to enhance discovery and personalization. Shaped's team will form Whatnot's applied AI research group, led by Murrell, to accelerate AI innovation across the marketplace.

How This Acquisition Fits into Whatnot's Broader AI Strategy

AI has been at the core of Whatnot since its founding. It already powers various aspects of the marketplace, from helping sellers create listings faster to improving trust and safety. Every week, Whatnot's systems process over half a million hours of live video and millions of real-time interactions, making recommendations smarter as the marketplace evolves. The addition of Shaped builds on this foundation, bringing complementary strengths across the AI stack—from infrastructure and research to production systems. As Whatnot's blog states, "Together, we're bringing complementary strengths across the AI stack, from infrastructure and research to production systems, so we can advance recommendations and match buyers to the right shows even faster."

The Challenges of Live Shopping Recommendations (and How Shaped Helps)

Live shopping presents unique challenges for recommendation systems. Unlike static e-commerce, where product catalogs remain relatively stable, live shopping involves dynamic inventories, unpredictable show durations, and real-time interactions that constantly shift what should be recommended next. Shaped's expertise in real-time ranking and search is tailor-made for this environment. As Shaped's blog explains, "When inventory changes minute to minute, shows run for hours, and every interaction shifts what should be ranked next, you can't just fine-tune a static retrieval and ranking stack. You have to rethink evaluation, exploration, and how you balance a buyer's next purchase against the long-term health of the marketplace." This acquisition allows Whatnot to tackle these challenges head-on.

What This Means for Sellers and Buyers

For sellers, faster, more accurate recommendations mean their products are more likely to be discovered by the right buyers at the right time. For buyers, it means a more personalized and engaging shopping experience, with recommendations that adapt to their interests and the live context. As Verrilli puts it, "You need to be recommending to folks based on what's in the show today and what that seller is selling today. And ultimately, buyers and sellers choose Whatnot because we can best match the two of them together." This acquisition enhances that matching capability, potentially leading to higher satisfaction and sales.

Shaped's Journey: From Startup to Acquisition

Shaped's path to acquisition is a testament to its innovative technology and team. The company kept encountering problems that standard recommendation playbooks couldn't solve, particularly in live marketplaces. As they noted, "We kept running into a category of problem that the standard recommendation playbook doesn't solve: live marketplaces." Whatnot, with its scale and data—processing over 500,000 hours of live video weekly and seeing cross-category buying up 170% year over year—presented the perfect opportunity to work on this problem directly. Rather than serving the problem from the outside, Shaped decided to join Whatnot and work on it at the company defining the future of live commerce.

The Future of Live Commerce: AI-Driven Discovery and Personalization

This acquisition signals Whatnot's commitment to leading the live commerce space through AI innovation. By integrating Shaped's real-time recommendation technology, Whatnot aims to make discovery more intuitive and personalized, helping users find new sellers, categories, and communities they might not have known they wanted. As Whatnot's blog emphasizes, "People might come looking for one thing, but they often leave having found a new seller, category, or community. Those unexpected discoveries are what power live shopping, and that's becoming even more important as the marketplace grows." With Shaped on board, Whatnot is poised to accelerate these discoveries, making live shopping even more engaging and efficient for everyone involved.

Find your next page

Tab to select · Enter to open · Esc to close