Summary
Divyam Goel is an MS Robotics student at Carnegie Mellon University with 8 years of research and industry experience building geometry-aware perception and generative models for robotics, vision, and speech. He develops adversarial, geometry-first testing frameworks for robot manipulation and integrates offline RL to improve high-speed grasping—work that earned an oral CoRL 2025 presentation under Prof. David Held and Prof. Zackory Erickson. Prior roles span GenAI model development at Adobe’s Firefly team and leading diffusion- and transformer-based projects for 3D scene generation, video avatars, and zero-shot TTS at research labs and startups. He has a strong systems-to-theory blend—shipping scalable cloud pipelines that produced 100k+ personalized videos while also publishing contributions in embodied AI, program comprehension, and generative modeling. Notably, he brings cross-domain expertise from efficient code-modeling and dataset creation to ergonomic-aware scene synthesis, allowing him to bridge practical production constraints with cutting-edge research.
8 years of coding experience
3 years of employment as a software developer
High School Science, High School Science at Delhi Public School Gurgaon
Indian Institute of Technology Roorkee
Master of Science - MS Robotics, Master of Science - MS Robotics at Carnegie Mellon University