Adi Renduchintala
Applied Research Scientist at NVIDIA
Redwood City, California, United States
Summary
🤩
Rockstar🎓
Top SchoolAdi Renduchintala is an applied research scientist with 12 years of experience building and deploying NLP, NMT, and large language model solutions, currently contributing to NVIDIA’s NeMo framework on adapter tuning for GPT and T5 models. He brings a strong research-to-production track record from roles at Meta and a PhD from Johns Hopkins, with deep expertise in machine translation, conversational AI, and learner-centric language modeling. Adi’s contributions include implementing IA3 adapter training, parallel adapter configurations, and bug fixes to enable scalable adapter integration in Megatron GPT, reflecting hands-on systems engineering as well as algorithmic skill. Earlier work spans industry-facing projects at Duolingo, Rosetta Stone, and clinical NLP, showing a recurring focus on robust language understanding for real users. Based in Redwood City, he blends academic rigor with practical ML engineering to move research innovations into widely used open-source tooling.
12 years of coding experience
9 years of employment as a software developer
M.S Arts Media and Engineering, M.S Arts Media and Engineering at Arizona State UniversityJohns Hopkins UniversityM.S Computer Science, M.S Computer Science at University of Colorado BoulderBE Electrical Engineering, BE Electrical Engineering at Anna University Chennai
Hindi
Github Skills (10)
neural-network10
transformer-models10
pytorch10
deep-learning10
large-language-models10
python10
generative-ai10
freetts4
gtts4
asr4
Programming languages (3)
Github contributions (5)
NVIDIA/NeMo
Jul 2022 - Jan 2023
A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)
Role in this project:
ML Engineer
ML EngineerContributions:233 reviews, 214 commits, 208 PRs in 6 months
Contributions summary:Adi contributed to adapter tuning for Megatron GPT models, specifically focusing on modifying the code for adapter learning and integrating adapters into the model architecture. They implemented and updated configurations, including parallel adapter configurations and IA3 support, for T5 and GPT models. The user's work involved the development of IA3 adapter training in GPT and T5 models, addressing reshaping for non-contiguous tensors, and fixing bugs.
asrspeech-recognitionnatural-language-processingttsspeaker-diarization
arendu/NeuralMachineTranslation_Tutorial_with_MxNet
Jun 2018 - Jan 2021
Contributions:15 commits in 2 years 7 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.