Data Infra for Embodied AI Happy Hour @ GTC
Data Infra for Embodied AI Happy Hour @ GTC
Join us during NVIDIA GTC for a relaxed evening with founders, engineers, researchers, and investors working on embodied AI, robotics, world models, and real-world data systems.
As embodied AI systems move from research into real-world deployment, the field is rapidly evolving across world modeling, data collection, policy learning, simulation, and system integration. This gathering brings together practitioners in robotics, autonomy, and multimodal AI to exchange perspectives on the technical roadmap for building intelligent systems that can perceive, reason, and act in the physical world.
We’ll kick off the evening with food and drinks after a full day at GTC, followed by a relaxed, no-slides conversation with startup founders, investors, and academics on the evolving embodied AI stack — from data pipelines and world models to training, deployment, and generalization challenges 🍻
Afterwards, the night opens up for casual networking and conversation with others building at the frontier of physical AI.
Schedule
Doors open at 7:30 PM
8 PM - 8:30 PM: Intros & Q&A
8:30 PM - 10:30 PM: Open Networking
Bites, drinks, and good vibes provided.
Space is limited. RSVP required.
Hosts
Oceanveo builds rich real-world data for training embodied AI systems. The company produces massive-scale, human-collected and annotated datasets across diverse environments, scenarios, and edge cases used for training and evaluating robotics and physical AI models.
Neural Motion is building a universal cross-embodiment data representation layer for embodied AI, turning robotics data that's currently scattered across formats, coordinate frames, and form factors into a standardized, reusable, embodiment-agnostic layer that foundational-model teams can build on.
Alchedata builds agentic, evaluator-driven multimodal data pipelines for modern AI systems. The company develops infrastructure for collecting, generating, and structuring data across image, video, audio, and text used in model training and evaluation.