Shashank Verma
Senior Developer Advocate Engineer - Deep Learning
California, United States
Summary
🤩
Rockstar🎓
Top SchoolShashank Verma is a Senior Developer Advocate Engineer specializing in deep learning with nine years of experience bridging research, product engineering, and developer ecosystems at NVIDIA. He builds end-to-end ML demos, writes technical content, and performs competitive analysis to help teams optimize model training and deployment—often translating complex research into production-ready workflows and MLOps practices. His background as a systems software engineer gives him rare fluency across low-level drivers, embedded platforms, and high-performance GPU inference, which he leverages to improve DL product performance. As an active contributor to the pytorch/TensorRT space, he has implemented and benchmarked SSD object-detection demos and optimized FP16 compilation paths to showcase inference gains. Based in California, he combines an MS in Electrical Engineering (ML & Computer Vision) with hands-on safety and multimedia firmware experience, enabling pragmatic solutions that span silicon to service. Colleagues rely on him for clear technical guidance that balances research rigor with real-world deployability.
9 years of coding experience
4 years of employment as a software developer
BITS Pilani, Birla Institute of Technology and ScienceMS Electrical Engineering, Specialization in Machine Learning & Computer Vision, MS Electrical Engineering, Specialization in Machine Learning & Computer Vision at University of Wisconsin-Madison
Hindi, English
Github Skills (11)
tensorrt10
cuda10
object-detection10
pytorch10
machine-learning10
deeplearning-ai10
deep-learning10
python9
benchmarking9
benchmark9
nvidia8
Programming languages (5)
Github contributions (5)
pytorch/TensorRT
Aug 2020 - Aug 2020
PyTorch/TorchScript/FX compiler for NVIDIA GPUs using TensorRT
Role in this project:
ML Engineer
ML EngineerContributions:10 commits, 4 PRs, 8 comments in 14 days
Contributions summary:Shashank primarily contributed to the development and demonstration of object detection models within the TRTorch framework. They added and refined an SSD (Single Shot MultiBox Detector) object detection demo notebook, showcasing the integration of TRTorch for performance optimization. Their work involved benchmarking the model, comparing performance before and after TRTorch integration, and compiling the model with FP16 precision. The user also addressed minor bugs and optimized code to achieve desired results.
cudapytorchjetsonnvidiadeep-learning
shashank3959/Network-Similarity
Mar 2020 - May 2020
Contributions:36 commits, 35 pushes, 1 branch in 2 months
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.