Sneha Mehta
Research Engineer at Meta
San Francisco Bay Area United States
Summary
👤
Senior🎓
Top SchoolSneha Mehta is a research scientist and NLP-focused AI/ML engineer with 13 years of experience building production-grade conversational AI and billion-scale recommender systems. Based in the San Francisco Bay Area, she has driven applied research and engineering across Meta, TikTok search and recommendations, Twitter Cortex research, and BloombergGPT post-training efforts. Her background spans both deep research—PhD-level training and work on knowledge graphs and recommender architectures—and hands-on productionization of large models and search relevance. She has practical experience in localization and machine translation from internships at Netflix and early R&D at DreamWorks, reflecting a breadth from creative pipelines to large-scale ML systems. Known for bridging research and product, she often moves prototypes into live serving environments at scale. A detail that sets her apart is combining formal math training in algebra and number theory with deep-learning specialization, informing rigorous model design and evaluation.
13 years of coding experience
7 years of employment as a software developer
Deep Learning Reinforcement Learning Summer School, Deep Learning Reinforcement Learning Summer School at University of AlbertaBITS Pilani, Birla Institute of Technology and ScienceDoctor of Philosophy (Ph.D.), Doctor of Philosophy (Ph.D.) at Virginia TechAlgebra and Number Theory, Algebra and Number Theory at Harish-Chandra Research Institute
English
Github Skills (13)
comprehension10
sentence10
nlp-machine-learning10
natural-language-processing9
nlp9
tensorboard6
visualization6
vector-space6
machine-learning5
pytorch5
chainer4
numpy3
mxnet2
Programming languages (1)
Github contributions (5)
sumehta/ideal_topic_modeling
Jan 2016 - Dec 2017
Contributions:8 commits, 7 pushes, 1 branch in 1 year 10 months
sumehta/FBMA
Jan 2020 - Oct 2020
Contributions:22 commits, 20 pushes, 1 branch in 8 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.