Vineeth Velmurugan

Vineeth is Encord's Robot Learning Lead. He joined to build Encord's internal data-creation team from the ground up, leading the effort to manufacture the real-world training data that physical AI models need, but that doesn't exist yet, from egocentric video and teleoperation logs to emerging modalities like forearm EMG sensors and brain-wave signals.

Before Encord, Vineeth worked on OpenAI's robotics team and served as Robotics Engineering Manager at OSARO, following an engineering management role at RIOS Intelligent Machines and nearly four years in robotics development at Berkshire Grey, the warehouse automation pioneer. That path, from warehouse automation to humanoid foundation models, shapes how he thinks about the physical AI data bottleneck: the constraint isn't model architecture, it's the scarcity of real-world interaction data.

Vineeth holds a degree from the University of Massachusetts Lowell.

Featured in:

• TechCrunch: Are brain waves the next unlock for physical AI? (Jul 26, 2026)
• Mezha: Encord creates real-world datasets to train physical AI (Jul 27, 2026)