Today’s websites, from e-commerce platforms to social networks, increasingly rely on recommendation engines to tailor their content for individual users. Shaped is stepping into the limelight with an $8 million Series A investment, offering businesses the tools to harness their existing data and integrate it with advanced language models and recommendation technologies for a customized user journey.
Shaped distinguishes itself as a platform built for developers, delivering unparalleled adaptability. Its toolkit encompasses a variety of data sources, ways of integration, models like Llama, CLIP, BERT, and methods for ranking recommendations and search results.

The minds behind Shaped include CEO Tullie Murrell, CPO Daniel Camilleri, and Shai Bruhis, who has ventured into new projects. Murrell’s expertise comes from his time at Meta/Facebook, particularly his involvement with PyTorchVideo and Facebook AI Research’s video AI endeavors. Camilleri brings experience from Uber and Afterpay. The duo initially crossed paths during a hackathon in their college days in Australia.
“Our core mission revolves around empowering smaller businesses to achieve the level of user personalization that’s currently seen at major tech firms, without the need for a large team of machine learning experts,” Murrell shared regarding Shaped’s foundational goals.
Among Shaped’s clientele are names like the adventure-driven Outdoorsy, Brazilian grocery app Trela, and QVC, which utilizes Shaped for generating live video recommendations on its app.
Even before the surge in generative AI technologies, Shaped was already incorporating state-of-the-art AI methodologies, a testament to Murrell’s background in AI research.
Initially focusing mainly on video personalization, Shaped, after participating in the Y Combinator accelerator in 2022, broadened its spectrum. “After our experiences in [YC], engaging with clients, and analyzing the market, we recognized the broader potential. We pivoted towards offering a versatile personalization platform for various types of media, including text, video, and audio,” commented Murrell.
Currently, Shaped can integrate data from a multitude of repositories, from data storage solutions like Databricks, Google’s BigQuery, and Snowflake, to PostgreSQL, MySQL, and MongoDB databases, alongside user engagement trackers like Amplitude, Google Analytics, and others. This extensive data collection, especially user interaction records, is crucial for developers creating tailored recommendation engines.
Focusing primarily on the development experience, Shaped prioritizes offering programmers the resources and data necessary for crafting and evaluating their systems. While being a code-first service, it also provides a dashboard for testing models and understanding the logic behind specific recommendations.
Lately, Shaped has delved deeper into enhancing its search capabilities. “With our deep semantic insights into user and content relationships, enhancing our search features felt like a natural evolution. This is a major focus for us moving forward, as we aim to develop a comprehensive discovery platform,” Murrell revealed.
The Series A round of Shaped was spearheaded by Madrona Ventures, accompanied by contributions from Y Combinator and leaders from notable firms such as Clickhouse, Docusign, and StitchFix.
“After meeting Shaped’s CEO, Tullie Murrell, over a year ago, we were thoroughly impressed with his deep knowledge of recommendation systems, gained through extensive research and development at Meta,” noted Madrona’s Karan Mehandru and Sabrina Wu in a blog post. “His clear vision, empathy for users, and insight into the market make him an ideal match for this venture. It’s also been a pleasure getting to know Dan, Ben, and the rest of the Shaped team, who bring rich experiences from their tenures at Uber, Apple, and other influential companies.”
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence


