Sami Abuelhaija
Co-Founder at Stealth Startup
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolSami Abuelhaija is a research-driven machine learning leader and entrepreneur with eight years of experience bridging Google Research and startup life; he most recently co-founded a stealth startup after serving as a Staff Research Scientist at Google. He combines deep academic training (PhD, USC) and practical ML engineering—contributing to high-profile open-source projects like YouTube-8M and TensorFlow GNN, where his work enabled smoother model inference, deployment, and TPU-friendly graph neural network training. His background spans end-to-end systems from static timing analysis and large-scale data pipelines to medical imaging research, giving him a rare full-stack perspective on ML research and production. Based in San Francisco, he is comfortable moving models from research prototypes to deployable artifacts (e.g., .tgz/.zip-ready models and Kaggle-compatible pipelines), and he brings founders’ pragmatism to technically ambitious problems.
8 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of Southern CaliforniaB.A.Sc. Honors, Electrical and Computer Engineering, B.A.Sc. Honors, Electrical and Computer Engineering at University of TorontoHigh school, High school at College De La SalleMaster of Science - MS, Computer Science and Engineering, Master of Science - MS, Computer Science and Engineering at University of MichiganSCPD (Non-degree option), Machine Learning, Computer Science, SCPD (Non-degree option), Machine Learning, Computer Science at Stanford University
Stackoverflow
1,989reputation
117kreached
5answers
2questions
Github Skills (24)
python10
gnn10
machine-learning10
inference10
tensorflow10
ai10
computer-vision10
graph-neural-network10
data-science9
continuous-deployment9
ml-deployment9
kaggle8
data-structures8
apache-beam8
data-structure8
Programming languages (2)
Github contributions (5)
google/youtube-8m
Oct 2017 - Jun 2018
Starter code for working with the YouTube-8M dataset.
Role in this project:
ML Engineer
ML EngineerContributions:38 commits, 10 PRs, 12 pushes in 8 months
Contributions summary:Sami made several significant contributions to the `youtube-8m` repository, primarily focused on improving the inference process and preparing the model for deployment. These contributions include updating inference and reader components to align with new datasets and Kaggle requirements. The user also added the ability to zip model files for easier sharing and integrated the model with .tgz files, streamlining model distribution and usage. In addition, the user updated code to fix the submission format to comply with Kaggle's requirements.
deep-learningdatasetmachine-learningyoutubeyoutube-8m
tensorflow/gnn
Jul 2022 - Dec 2022
TensorFlow GNN is a library to build Graph Neural Networks on the TensorFlow platform.
Role in this project:ML Engineer
ML EngineerContributions:30 commits, 1 comment in 5 months
Contributions summary:Sami contributed to the development of graph neural networks (GNNs) within the TensorFlow GNN library. Their work included implementing support for Planetoid datasets (Cora, Citeseer, Pubmed) for in-memory training, showcasing node-wise transformations, and implementing the GCN model as an example. They further added features like the ability to train on train+validation splits and integer mathematics for self-loops to optimize TPU-friendliness.
gnndeep-learningneural-graphneural-networksmachine-learning
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.