Arvind Sridhar is an accomplished ML-focused software engineer and researcher with over a decade of experience building and optimizing production-grade AI systems and perception pipelines for autonomous vehicles and large-scale models. He has transitioned fluidly between industry and academia—contributing to Google, NVIDIA, Waymo, Argo AI and Stanford—where his work spanned multimodal foundation model tuning, efficient Transformer and Graphormer architectures, and certified robustness for vision models. Arvind combines deep systems know-how (notably contributions to the pytorch/tensorrt compiler to accelerate PyTorch→TensorRT deployments) with hands-on product and internship experience at Uber ATG, Nuro, and Citadel. He’s taught top-tier ML and graph/LLM courses, published first-author research at ICLR/ICML workshops, and has a track record of shipping low-latency inference solutions for real-time applications. Based in Sunnyvale, he blends rigorous research instincts with pragmatic engineering to move models from lab to edge.
11 years of coding experience
9 years of employment as a software developer
Bachelor of Science (BS), Electrical Engineering and Computer Science, Bachelor of Science (BS), Electrical Engineering and Computer Science at University of California, Berkeley
High School, High School at Bellarmine College Preparatory
Bachelor of Science (BS), Economics, Finance, and Business Administration, Bachelor of Science (BS), Economics, Finance, and Business Administration at University of California, Berkeley, Haas School of Business
Master of Science (MS), Computer Science and Artificial Intelligence, Master of Science (MS), Computer Science and Artificial Intelligence at Stanford University
PyTorch/TorchScript/FX compiler for NVIDIA GPUs using TensorRT
Role in this project:
ML Engineer
Contributions:1 review, 15 commits, 3 PRs in 2 months
Contributions summary:Arvind primarily contributes to the `pytorch/tensorrt` repository by implementing and testing functionality related to converting PyTorch models to TensorRT for NVIDIA GPUs. The commits focus on bug fixes and adding tests for specific operators, such as `gru_cell` and `lstm_cell`, ensuring correct conversion and functionality within the TensorRT framework. Code changes include modifications to the converter implementations and the addition of test cases to validate the converted models.
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.