Abhijeet Awasthi
Applied Scientist at Amazon Web Services (AWS)
Menlo Park, California, United States
Summary
👤
Senior🎓
Top SchoolAbhijeet Awasthi is an applied scientist with nine years of experience specializing in data-efficient training and adaptation of large sequence models across text, speech, and code. He holds a PhD from IIT Bombay and has a strong publication record in top ML and NLP venues (ICML, ICLR, ACL, EMNLP, INTERSPEECH), with research spanning few-shot learning, synthetic data strategies, and architecture design for low-label regimes. His industry work includes LLMs for code and copilots at Microsoft, multimodal embedding research at Meta, and current applied science on Code-LLMs and agentic RL at AWS, reflecting a rare blend of academic rigor and product-minded impact. Notably, his collaborations with Google Research during his PhD produced multilingual and speech systems that scaled to dozens of languages, demonstrating a practical focus on real-world deployment and customization.
9 years of coding experience
6 years of employment as a software developer
High School, High School at New Digamber Public School, IndoreSenior Secondary School, Senior Secondary School at Daisy Dales School, IndoreIndian Institute of Technology BombayB.TECH. Electronics and Electrical Communication Engineering, B.TECH. Electronics and Electrical Communication Engineering at Indian Institute of Technology, Kharagpur
Github Skills (88)
python10
deep-learning10
representation-learning10
pre-trained-model10
bert-embeddings10
sequence-labeling10
natural-language-understanding10
tensorflow10
google10
weakly-supervised-learning10
error-correction10
machine-learning10
bert10
machine-translation9
ml9
Programming languages (5)
Github contributions (5)
awasthiabhijeet/refill
Nov 2022 - Dec 2022
Diverse Parallel Data Synthesis for Cross-Database Adaptation of Text-to-SQL Parsers (EMNLP 2022)
Contributions:4 commits, 3 pushes, 1 branch in 1 month
cross-databasesql-parser
awasthiabhijeet/PIE
Oct 2019 - Sep 2021
Fast + Non-Autoregressive Grammatical Error Correction using BERT. Code and Pre-trained models for paper "Parallel Iterative Edit Models for Local Sequence Transduction": www.aclweb.org/anthology/D19-1435.pdf (EMNLP-IJCNLP 2019)
Contributions:22 commits, 1 PR, 17 pushes in 1 year 11 months
berterror-correctionpre-trained-modelgrammatical-error-correctionsequence-transduction
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.