Sami Wilf
United States
Summary
🤩
Rockstar🎓
Top SchoolSami Wilf is an AI systems engineer with seven years of experience building production-ready machine learning infrastructure and leading engineering teams. He has held technical and leadership roles from software engineer to VP of Engineering, including contributions at Meta on AI system SW/HW co-design and multi-year leadership at IntelliScience. Sami blends deep ML engineering—evidenced by contributions to the widely used PyTorch torchrec recommendation library, improving data-loading and multi-node training robustness—with practical product delivery and consulting experience. He excels at bridging research, low-level system optimizations, and scalable deployment, and often focuses on edge cases and testing to harden models in real-world settings. Based in the United States, he pairs a CS degree from Georgia State with a track record of shipping reliable, high-performance AI systems.
7 years of coding experience
12 years of employment as a software developer
BS, Computer Science, BS, Computer Science at Georgia State University
Github Skills (14)
mle10
pytorch10
deeplearning-ai10
recommendation-system10
deep-learning10
python10
ml10
numpy9
machine-learning9
distributed-training9
cuda8
testing8
data-pipeline8
data-pipelines8
Programming languages (4)
Github contributions (5)
pytorch/torchrec
Mar 2022 - Dec 2022
Pytorch domain library for recommendation systems
Role in this project:
ML Engineer
ML EngineerContributions:1 review, 9 commits, 28 PRs in 8 months
Contributions summary:Sami contributed to the optimization and enhancement of the PyTorch-based recommendation systems library. They focused on improving data loading efficiency by implementing `mmap_mode` for the Criteo dataset and addressing issues with multi-node training, ensuring correct device assignments. They also made changes to the DLRM model to handle potential edge cases and added comprehensive unit tests for the Criteo dataset, validating the validation and test sets.
cudapytorchrecommendation-systemsdeep-learninggpu
facebookresearch/FAMBench
Oct 2021 - Jan 2023
Benchmarks to capture important workloads.
Contributions:8 reviews, 53 commits, 36 PRs in 1 year 3 months
stress-testingbenchmarkingworkloadsbenchmarkbenchmarks
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.