Siddharth Patki is a Principal Research Scientist with nine years of experience building foundation models and applied AI for robot manipulation, navigation, and human-robot collaboration. He has driven multi-modal LLM post-training work at Meta (including core contributions to Llama4 post-training) and translated research into production-ready robotics solutions at Amazon and General Motors. His PhD and postdoc focused on language-grounded perception and multi-agent task planning, producing open-source tools like the H2SL ROS package and multiple top-tier publications. Now at NVIDIA, he focuses on foundation models for Physical AI, blending deep learning, simulation-driven datasets, and RL/SFT techniques to close real-world model gaps. Colleagues note his uncommon ability to bridge rigorous academic research with fast-paced, product-oriented model engineering across perception, planning, and multi-modal LLM stacks.
9 years of coding experience
10 years of employment as a software developer
Bachelor of Technology (B.Tech.) Electronics Engineering, Bachelor of Technology (B.Tech.) Electronics Engineering at Vishwakarma Institute Of Technology
Doctor of Philosophy - PhD Electrical Engineering, Doctor of Philosophy - PhD Electrical Engineering at University of Rochester
Download and visualize single or multiple classes from the huge Open Images v4 dataset
Contributions:9 pushes in 2 years 2 months
visualizeopen-imagespythondataset
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.
Request Free Trial
Siddharth Patki - AI Research Scientist at General Motors