Memories.ai Secures $16M to Build Visual Memory Layer
Memories.ai, an artificial intelligence startup founded in 2024 by former Meta AI team members Shawn Shen and Ben Zhou, has completed a $16 million financing round. The funding was led by Susa Ventures and included contributions from Seedcamp, Fusion Fund, and Crane Venture Partners. The capital was raised through an initial $8 million seed round in July 2025, followed by an $8 million extension, signaling strong investor confidence in the company's specialized focus on visual AI memory.
The founders, whose experience includes working on the AI system for Meta's RayBan glasses, identified a critical gap in the market. While major players like OpenAI and Google have developed text-based memory for their models, Memories.ai is building the foundational infrastructure to allow AI in physical devices to remember, index, and recall visual data. This capability is essential for the next generation of smart wearables and autonomous robots.
Nvidia Partnership Validates Visual-First AI Strategy
In a key strategic move announced at Nvidia's GTC conference, Memories.ai has partnered with the semiconductor giant. The collaboration will see Memories.ai leverage Nvidia's advanced AI tools, including the Cosmos-Reason 2 reasoning vision language model and the Metropolis video search platform. This partnership provides Memories.ai with powerful technology and validates its approach to solving the complex challenge of unstructured visual data.
Unlike text, which is structured and easier to index, video and real-world visual inputs require sophisticated models to be useful. Shen and Zhou's insight from their time at Meta was that without the ability to recall what they see, AI-powered glasses and other devices would have limited real-world utility. Their company aims to provide this "visual memory," which has been a secondary focus for most large AI labs.
Infrastructure-First Approach Targets Future Wearables Market
Memories.ai is concentrating its efforts on developing its Large Visual Memory Model (LVMM) and the supporting infrastructure, rather than consumer hardware. While the company developed its own 'LUCI' recording device for data collection, it has no plans to become a hardware manufacturer. This focus on core technology is further reinforced by a separate partnership with Qualcomm to ensure its models can run efficiently on its processors.
The company is already engaged with large wearable companies, though it has not disclosed names. Co-founder Shawn Shen has indicated a long-term commercialization strategy, acknowledging the market's nascent stage.
In terms of commercialization, we are more focused on the model and the infrastructure, because ultimately we think the wearables and robotics market will come, but it’s probably just not now.
— Shawn Shen, Co-founder, Memories.ai