Abheesht Sharma
Machine Learning Engineer at Google
Bengaluru, Karnataka, India
Summary
🤩
Rockstar🎓
Top SchoolAbheesht Sharma is a Machine Learning Engineer at Google working on Keras/JAX where he architects multi-backend, pretrained model tooling like KerasHub and practical libraries for recommender systems and post-training LLM workflows. With six years of experience spanning applied science and research roles at Amazon and top academic collaborations, he has productionized billion-scale fraud-detection models and contributed core NLP models and APIs to KerasNLP. His open-source impact includes substantive contributions to the widely used keras-hub (e.g., FNet encoder, perplexity metric, byte tokenizer) and recognition via Google AI/ML community awards. Comfortable moving models from research to realtime systems, he blends deep theoretical rigor (published at EMNLP/JMLR) with hands-on engineering for high-throughput environments. Outside work he fuels creativity through voracious reading and informal storytelling, a trait that surfaces in his lucid documentation and developer-focused tools.
6 years of coding experience
3 years of employment as a software developer
Bachelor's, Computer Science, 9.17, Bachelor's, Computer Science, 9.17 at Birla Institute of Technology and Science, Pilani - Goa Campus
Github Skills (5)
keras10
machine-learning10
nlp10
tensorflow10
unit-testing9
Programming languages (4)
Github contributions (5)
keras-team/keras-hub
Mar 2022 - Jan 2023
Pretrained model hub for Keras 3.
Role in this project:
ML Engineer
ML EngineerContributions:2 releases, 304 reviews, 69 commits in 10 months
Contributions summary:Abheesht made substantial contributions to the `keras-hub` repository, focusing on the implementation and integration of various machine learning models and components. This included the addition of an FNet encoder layer, along with its corresponding unit tests, showcasing expertise in creating new model architectures. Furthermore, the user introduced a perplexity metric, and contributed to the development of a Byte Tokenizer to enhance KerasNLP, demonstrating skills in model development and performance evaluation. The user also fixed numerous documentation issues.
nlppythondeep-learninglanguage-processingmachine-learning
abheesht17/Abstractive-Text-Summarisation
Apr 2020 - Mar 2025
Contributions:8 pushes, 1 branch in 4 years 11 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.