Moin Nadeem
Co-Founder at Phonic
California, United States
Summary
🤩
Rockstar🎓
Top SchoolMoin Nadeem is a Co-Founder and machine learning engineer based in California with 10 years of experience building production ML and voice-AI systems. He currently leads Phonic, a Reliable Voice AI startup, after starting the NLP team at MosaicML as the company's eighth employee and onboarding its first Fortune 500 customer. His research background spans MIT CSAIL and Stanford—work that includes autoregressive generation, factuality in LMs, and an unusual project turning prompts into websites. An active open-source contributor, he improved training infrastructure in MosaicML's widely used composer library (LR schedules, checkpointing, object-storage loading, and BERT support), and has consulted for LLM startups on price/performance inference. He blends academic rigor with hands-on product and customer-facing engineering.
10 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Stanford University
English, Spanish, Urdu, Arabic
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (8)
neural-network10
pytorch10
machine-learning10
deep-learning10
python10
ml10
pytorch-lightning9
bert8
Programming languages (15)
Github contributions (5)
mosaicml/composer
Oct 2021 - Aug 2022
Supercharge Your Model Training
Role in this project:
ML Engineer
ML EngineerContributions:202 reviews, 55 commits, 94 PRs in 10 months
Contributions summary:Moin primarily contributed to the implementation of training-related components within the `composer` library. Their commits involved adding and improving learning rate scheduling algorithms such as linear decay and cosine annealing, as well as refining checkpointing mechanisms. They also introduced support for loading checkpoints from object storage and incorporated BERT models, demonstrating a focus on enhancing training efficiency and expanding the library's capabilities for deep learning tasks.
pytorchml-systemsdeep-learningneural-networksmachine-learning
moinnadeem/composer
Mar 2022 - Sep 2022
library of algorithms to speed up neural network training
Contributions:45 PRs, 297 pushes, 69 branches in 5 months
speeddeep-learningneural-networksmachine-learningtraining
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.