Video AI for Real-World Systems Social

Join us for an evening hosted by a team building video AI systems for real-world deployment, bringing together researchers, engineers, and builders working at the frontier of video AI. This is a relaxed, conversation-first gathering connecting people across robotics, self-driving, and embodied AI.

We’ll focus on a new shift in how video is used in real-world AI: moving beyond static labeling toward systems that can search, reason over, and validate events directly from raw footage. Discussions will explore approaches that combine vision-language models, motion understanding, and multi-stage validation to turn large-scale video into something you can actually act on.

No formal talks, just high-signal conversations with people actively building and deploying these systems.

The goal is to go beyond surface-level ideas and get into what it actually takes to make Physical AI work reliably in the real world. If you are working on autonomy, VLMs, or production AI systems, this is a chance to exchange ideas and shape what comes next.