Paarth Neekhara
Research Scientist at NVIDIA
Uttarakhand, India
Summary
🤩
Rockstar🎓
Top SchoolPaarth Neekhara is a research scientist and PhD candidate in computer science at UCSD with 11 years of engineering experience bridging academic research and production software. He currently works on conversational AI and speech-language models at NVIDIA, after earlier research internships focused on voice conversion, speaker adaptation, and deepfake detection at leading labs including NVIDIA and Facebook AI. Prior to his research career he spent years as a technical lead and software architect building fleet- and fuel-optimization platforms for shipping at Blue Water Trade Winds, giving him strong systems and product instincts. He contributes to open-source ML projects—most notably a text-to-image synthesis repo where he designed data pipelines and model initialization for GAN-based generation—and often focuses on the messy, high-impact work of preprocessing and model bootstrapping. Combining rigorous academic training (PhD, UCSD; BTech, IIT Roorkee) with production-grade engineering, he is comfortable moving ideas from proofs-of-concept to deployed systems. Based in Uttarakhand, India, he blends deep learning research with real-world optimization problems across industry and cloud-scale environments.
11 years of coding experience
University of California, San DiegoIndian Institute of Technology Roorkee
Github Skills (7)
data-preprocessing10
machine-learning10
nlp10
tensorflow10
python10
natural-language-processing10
image-processing9
Programming languages (3)
Github contributions (5)
paarthneekhara/text-to-image
Aug 2016 - Jan 2018
Text to image synthesis using thought vectors
Role in this project:
ML Engineer
ML EngineerContributions:32 commits, 4 PRs, 13 pushes in 1 year 5 months
Contributions summary:Paarth primarily worked on data preprocessing and model initialization for a text-to-image synthesis project. Their commits involve loading and processing image captions, integrating word embeddings (GloVe), and preparing the data for model training. The code changes include modifications to data loading, image processing, and the initialization of a generative model (GAN). This suggests the user is focused on the data pipeline and the initial setup of the machine learning model.
image-synthesisvectorssynthesisthoughttext-to-image
paarthneekhara/AdversarialDeepFakes
Nov 2019 - Feb 2022
Contributions:65 commits, 1 push in 2 years 3 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.