We are building the visual memory infrastructure for AI. The layer that lets robots, wearables, and agents hold on to what they see, so the more they operate, the more they understand.
Founded to solve a problem every video-heavy enterprise faces.
Mountains of footage, and almost none of it usable. Memories.ai started with that observation and a thesis that visual memory deserves its own field. Three years later we are deployed across physical AI, media, security, and consumer devices, with partners that include NVIDIA, Lenovo, Samsung, Vivo, Ring, and Comcast.
The Team
Shawn Shen
Founder & CEO
University of Cambridge, PhD, Research Scientist
Ben Zhou
Co-Founder & CTO
ML Engineer, Lead in On-Device Perception
Eddy Wu
Chief AI Officer
Principal Research Scientist, Meta Superintelligence Lab, Lead in Multi-Modal AI
Ryan Gaertner
Chief Operating Officer
Prev. Founder/CEO & Investor, Stanford MBA '23 & BS’17, Stanford Football ’17
Advisors
"Memories.ai gave our partners a shared retrieval layer that holds up across regions. What used to take weeks of pipeline work now happens in hours — and the lineage is built in."
Dr. Lena Park
Director of Research Infrastructure, NVIDIA Partner Program
What we care about
Vision — why do we exist?
To be the visual memories for every physical intelligence in the world. As physical intelligence moves from labs into the real world, the bottleneck won't be actuation or control — it will be memories.
Mission — what are we setting out to accomplish?
To let physical intelligence self-evolve through visual memories. We turn what it sees into persistent, structured memories it can learn from — so every embodied agent gets better with experience, and the whole fleet gets smarter together.
Values — how we work, what we value
Build good memories. The memories we build for physical intelligence must be real, reliable, and worth trusting — and the memories we make as a team should be ones we're proud of.