Rajeev Rao
Software Engineering Manager at NVIDIA
California, United States
Summary
🤩
Rockstar🎓
Top SchoolRajeev Rao is a Software Engineering Manager at NVIDIA with 11 years of experience building and shipping high-performance machine learning systems. He leads teams focused on ML software and inference optimization, contributing hands-on to flagship open-source projects like NVIDIA DeepLearningExamples and the TensorRT SDK. His work includes implementing custom TensorRT plugins (including stable diffusion support), optimizing CUDA initialization and inference pipelines, and improving reproducibility and performance for enterprise deployments. Based in California, he combines an MS in Electrical and Computer Engineering from UC Santa Barbara with a background in electronics engineering to bridge research-quality models and production-grade GPU inference. Notably, he blends managerial leadership with active low-level engineering, ensuring teams deliver both developer-friendly samples and highly optimized inference components.
10 years of coding experience
BE, Electronics and Communication Engineering, BE, Electronics and Communication Engineering at Ramaiah Institute Of TechnologyMS, Electrical and Computer Engineering, MS, Electrical and Computer Engineering at University of California, Santa Barbara
English
Github Skills (20)
c-language10
tensorrt10
plugin-development10
machine-learning10
inference10
mlops10
deep-learning10
optimisation10
bert10
cuda10
cprogramming-language10
optimization10
pytorch9
onnx9
computer-engineering8
Programming languages (5)
Github contributions (5)
NVIDIA/TensorRT
Jun 2019 - Dec 2022
NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:34 releases, 87 reviews, 254 commits in 3 years 6 months
Contributions summary:Rajeev's commits reveal involvement in developing and maintaining core components of NVIDIA's TensorRT SDK. The primary focus is on the implementation of custom plugins for the efficient inference of machine learning models. The user has specifically added support for, and implemented, custom TensorRT plugins for stable diffusion, indicating a deep understanding of the TensorRT ecosystem. Furthermore, the commits include contributions to documentation and sample code, demonstrating a dedication to making the library accessible and user-friendly.
deep-learning-inferencenvidiadeep-learningruntimeinference
NVIDIA/DeepLearningExamples
Dec 2019 - Sep 2020
State-of-the-Art Deep Learning scripts organized by models - easy to train and deploy with reproducible accuracy and performance on enterprise-grade infrastructure.
Role in this project:MLOps Engineer
MLOps EngineerContributions:12 commits, 5 PRs, 1 comment in 9 months
Contributions summary:Rajeev primarily contributed to integrating and optimizing deep learning models within a TensorRT environment. Their work involved initializing CUDA states, integrating BERT demos, and fixing regressions related to TensorRT-7.0. The user also updated the Jasper sample to TensorRT 7.1.3.4 and addressed review comments across various files. This indicates a focus on deploying and optimizing models for efficient inference.
forecastingcaffe2translationspeech-recognitionstate-of-the-art
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.