Abhinav Arora
Staff Research Engineer at Google
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolAbhinav Arora is a Staff Research Engineer in the San Francisco Bay Area with 11 years of software engineering experience and over 8 years focused on machine learning and deep learning, including four years leading cross-functional teams of engineers, computational linguists, and scientists. Currently driving post-training research to improve multi-turn dialogue, groundedness, and search behaviors for Gemini at Google, he previously led ML engineering efforts at Meta and contributed production-quality ML tooling at Baidu and Adobe. He is an active open-source contributor to major projects like PaddlePaddle and torchtext, with hands-on work that ranges from deep framework operator fixes to Torchscriptable tokenizers and vocabulary tooling. Known for improving code quality and deployability, he blends research rigor with pragmatic backend engineering to ship robust, scalable ML systems.
11 years of coding experience
12 years of employment as a software developer
B.E, Information Technology, B.E, Information Technology at Netaji Subhas Institute of TechnologyClass 10, Class 10, Class 10, Class 10 at Air Force Golden Jubilee InstituteMaster of Computational Data Science, Analytics, Master of Computational Data Science, Analytics at Carnegie Mellon UniversityClass 12, Class 12, Class 12, Class 12 at Ryans International School
English, Hindi, Sanskrit
Github Skills (19)
torchscript10
tokenize10
pytorch10
c-language10
python10
machine-learning10
tokenizer10
deep-learning10
text-processing10
byte-pair-encoding10
paddlepaddle10
gpt10
nlp10
cprogramming-language10
computer-engineering9
Programming languages (4)
Github contributions (5)
facebookresearch/pytext
Mar 2019 - Mar 2022
A natural language modeling framework based on PyTorch
Role in this project:
Back-end Developer
Back-end DeveloperContributions:12 commits, 6 PRs in 3 years
Contributions summary:Abhinav primarily focused on enhancing the PyText framework, which is a natural language modeling framework based on PyTorch. Their contributions include fixing bugs in dictionary field padding at the minibatch level, adding support for TorchScript export of intent slot output layers, CRF, and XLM intent slot models. Additionally, they worked on torchscriptifying the `RobertaSeqLabelingtensorizer` and related components to enhance model compatibility with TorchScript and improved the tokenization process.
pytorchnlpbertmachine-learningnatural-language
PaddlePaddle/Paddle
Dec 2017 - May 2018
PArallel Distributed Deep LEarning: Machine Learning Framework from Industrial Practice （『飞桨』核心框架，深度学习&机器学习高性能单机、分布式训练和跨平台部署）
Role in this project:Back-end Developer
Back-end DeveloperContributions:65 commits, 298 PRs, 145 pushes in 4 months
Contributions summary:Abhinav primarily contributed to fixing CPPLint errors across various operator files within the PaddlePaddle framework. They addressed code style issues and resolved specific errors in operators related to prior boxes, pooling, and softmax calculations. Additionally, the user corrected errors in send/receive and related utility files, demonstrating a focus on code quality and consistency within the deep learning framework.
pytorchpythonparalleldeep-learningpaddlepaddle
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.