Parth Radia Email & Phone Number

California, United States

California, United States

Los Angeles, California, United States

Secured Ericsson C-suite support to establish a new division focused on generative AI in the context of communication technologies. Built and now lead a team of 30 rockstar scientists, engineers and product people at the forefront of photorealistic avatars, lip sync, and voice/video synthesis. This pioneering work has required the collection of a petascale 5D human dataset in a state-of-the-art light stage, one of few such efforts in the world. Leveraged this foundational work to bring Ericsson's first generative AI product to market: deep voice and video translation (see linked videos). 100s of primarily enterprise partners and customers across the globe. Also manage the division’s P&L, overseeing annual budget to drive innovations and adoption throughout the funnel.

Palo Alto

Transitioned to leading a specialized, cross-functional team focused on bringing foundational research from my previous role (photorealistic digitization and animation of humans) to the real world application of 3D presence. Provided leadership and contributions across the stack, from AI and XR research to product design.In November 2020, our team achieved the world’s first* ‘zero-infrastructure’ photorealistic 3D call over production networks—requiring only a commodity HMD for each user. We then leveraged our tech to broadcast the world's first* XR concert via a single modern smartphone.*To the best of our knowledge.

Palo Alto, California, United States

Founded and lead research efforts in the photorealistic digitization and animation of humans for 3D presence. We built systems that discovered novel neural networks for accurately and speedily (60 FPS) driving photorealistic 3D meshes given the constraint of limited signal at inference time. (E.g. driving a photorealistic talking head with just sensory information from commodity headsets.) As part of this effort, raised funds internally to build a state-of-the-art light stage and deep learning pipeline.