Elad Eban
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolElad Eban is a machine learning researcher, tech lead and manager with seven years of experience bringing theory-driven ML research into production, currently based in the San Francisco Bay Area and working at Google. He combines rigorous academic training—a PhD and top honors in CS from The Hebrew University—with hands-on engineering, notably contributing to Google Research’s morph-net to optimize resource-constrained deep networks (adding Conv3D, ResizeBilinear, Conv2DTranspose support and refining cost/latency calculations). Elad thrives on fast-paced, creative teams solving open-ended, real-world problems and has entrepreneurial experience founding a consulting startup. Colleagues rely on him to bridge deep theory and pragmatic system design, turning research prototypes into deployable solutions. An interesting detail: he pairs strong publication-level research instincts with practical code-level improvements that materially reduce model latency and deployment cost.
7 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 The Hebrew University of JerusalemBachelor's Degree, Computer Science and Computational Biology, 96.5, Bachelor's Degree, Computer Science and Computational Biology, 96.5 at The Hebrew University
Hebrew, French, English
Github Skills (6)
machine-learning10
deep-learning10
tensorflow10
python10
neural-architecture-search10
automl8
Programming languages (3)
Github contributions (4)
google-research/morph-net
Mar 2019 - May 2020
Fast & Simple Resource-Constrained Learning of Deep Network Structure
Role in this project:
ML Engineer
ML EngineerContributions:26 commits, 1 PR, 1 push in 1 year 2 months
Contributions summary:Elad primarily contributed to the `morph-net` repository by modifying and extending the codebase to support the structure of deep neural networks. They added new features like support for Conv3D and ResizeBilinear operations and incorporated Conv2DTranspose into the structure exporting process. They also refined existing code, improving the readability and consistency of the regularizer and its integration into the network. The user's work involves changes related to cost and latency calculations which are core to this machine learning project.
pythondeep-learningneural-architecture-searchneural-networksmachine-learning
google-research/google-research
Jan 2021 - Jan 2021
Google Research
Contributions:1 commit, 2 comments in 1 day
googlemachine-learningai
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.