Aman Gupta
Software Engineer III at Google India Pvt. Ltd.
Bellevue, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolAman Gupta is a Software Engineer III at Google with 11 years of experience building high-performance C++ systems and machine learning tooling, including hands-on contributions to TensorFlow TPU and models repositories. He began his career developing image processing and OCR research in Indian languages and later shipped core C++ features for Photoshop Elements at Adobe. At Google he has bridged ML engineering and systems work—implementing custom TPUEstimator flows, distributed Mask R-CNN enhancements, and model serving improvements—demonstrating both low-level performance focus and production ML delivery. Comfortable across Linux and Windows toolchains, Aman combines classical signal/image processing expertise (MFCC, DTW, binarization, segmentation) with modern ML infrastructure experience, and he’s based in Bellevue, Washington. An intriguing throughline in his work is a consistent drive to move research-grade algorithms into robust, deployable code.
11 years of coding experience
5 years of employment as a software developer
Senior Secondary, English, Physics, Chemistry, Biology, Maths, Senior Secondary, English, Physics, Chemistry, Biology, Maths at Lord Mahavira AcademyB.Tech, Information Technology, B.Tech, Information Technology at Indian Institute of Information Technology, Allahabad
Stackoverflow
275reputation
13kreached
13answers
0questions
Github Skills (18)
python10
machine-learning10
tpu10
deep-learning10
tensorflow10
evaluation9
distributed-training9
trainings9
eval9
modeling9
computer-vision9
datasets8
google-compute-engine6
google-cloud-ml6
jupyter-notebook6
Programming languages (4)
Github contributions (5)
tensorflow/tpu
Oct 2018 - May 2022
Reference models and tools for Cloud TPUs.
Role in this project:
ML Engineer
ML EngineerContributions:64 commits, 49 PRs, 62 pushes in 3 years 7 months
Contributions summary:Aman merged internal changes into the public repository related to the `tpu` project, particularly focusing on a custom TPUEstimator implementation for the Iris dataset. The code changes involved the integration of data loading, model definition, training, and evaluation using TensorFlow, demonstrating a clear focus on utilizing Cloud TPUs. Further contributions included modifications to the Mask RCNN model, including distributed execution, evaluation metrics, and visualization tools. These changes also included model serving capabilities.
cloud
tensorflow/models
Aug 2018 - Nov 2021
Models and examples built with TensorFlow
Role in this project:ML Engineer
ML EngineerContributions:1 review, 10 commits, 2 PRs in 3 years 3 months
Contributions summary:Aman contributed to the TensorFlow models repository by modifying and enabling prediction functionality within the MNIST TPU example. They specifically addressed the lack of input data for prediction by using a subset of the test data and updated code related to the TPUEstimator for prediction. The user also addressed an issue in the AutoAugment training script involving a dataset download URL.
deep-learningtensorflow
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.