Harshitha Venkata
Software Engineer - Machine Learning at Meta
Menlo Park, California, United States
Summary
🤩
Rockstar🎓
Top SchoolHarshitha Venkata is a software engineer specializing in machine learning with six years of experience building and optimizing ML training and inference frameworks at Meta and Microsoft. She combines a strong applied mathematics background with hands-on NLP and computer vision work, having contributed to ML.NET image classification APIs and core ONNX Runtime gradient operations. Her experience spans research institutions and industry, from the University of Utah and IISc to production ML at PASSUR Aerospace and major cloud-scale projects. At Microsoft she implemented learning rate schedulers and optimizer integrations; at ONNX Runtime she debugged gradient ops and added Deepspeed/fairscale training samples, reflecting depth in scalable training tooling. Based in Menlo Park, she brings both research rigor and production engineering discipline to ML systems. An understated strength is her pattern of improving robustness—fixing tests, addressing feedback, and documenting defaults to make frameworks more reliable for downstream users.
6 years of coding experience
8 years of employment as a software developer
Bachelor's Degree, Electrical, Electronics and Communications Engineering, Bachelor's Degree, Electrical, Electronics and Communications Engineering at Visvesvaraya Technological UniversityThe University of Utah
English, Telugu, Kannada, Hindi
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (22)
pytorch10
mldotnet10
image-classification10
net10
machine-learning10
dotnet10
onnx10
asp-net10
deeplearning-ai10
deep-learning10
tensorflow10
mlnet10
dotnet-core10
computer-vision10
csharp10
Programming languages (4)
Github contributions (5)
microsoft/onnxruntime
Jul 2020 - Mar 2022
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:
ML Engineer
ML EngineerContributions:1 release, 30 reviews, 39 commits in 1 year 8 months
Contributions summary:Harshitha primarily contributes to the ONNX Runtime project by addressing issues related to the training of machine learning models. They focus on implementing and debugging specific gradient operations such as Tile, and Where, which are vital for the proper training of ONNX models. The user's work also includes fixing bugs, code cleanup, and bumping onnxruntime version, showcasing their active involvement in the project's maintenance and improvement. Moreover, they added samples for Deepspeed pipeline parallel and fairscale sharded optimizer with ortmodule, demonstrating integration of training optimizations.
runtimetrainingtensorflowai-frameworkaccelerator
dotnet/machinelearning
Aug 2019 - Dec 2019
ML.NET is an open source and cross-platform machine learning framework for .NET.
Role in this project:ML Engineer
ML EngineerContributions:11 commits, 19 PRs, 3 pushes in 3 months
Contributions summary:Harshitha primarily contributed to the ML.NET framework, focusing on image classification functionality. Their work involved modifying the image classification API to handle image data as `VBuffer<byte>`, fixing bugs in unit tests, and addressing feedback. They also implemented the `LearningRateScheduler` functionality, including decay methods, for the image classification trainer, which included adding the GradientDescentOptimizer. Further contributions involved documentation updates and changes to default settings within the image classification API.
machine-learning-platformdotnetml-netmachine-learningcsharp
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.