Jakub Tomczak
Senior Staff Research Scientist at NatInLab B.V.
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolJakub Tomczak is a Senior Staff Research Scientist and Generative AI leader with 15+ years of machine learning and deep learning experience and nine years in industry roles spanning research, productization, and leadership. He has led large academic and industrial programs—supervising 50+ MSc/PhD students, managing cross-functional teams, securing EUR 2.3M in funding, and serving as Program Chair for NeurIPS 2024. Jakub combines hands-on model engineering (notably deep generative models implemented in his intro_dgm repository) with strategic fractional CTO/advisory roles for companies including eBay and Qualcomm. His work spans foundation models, LLMs, computer vision, and GenAI for drug discovery, yielding 2 patents and 20+ top-conference papers. Comfortable switching between code, grants, and executive strategy, he’s known for fostering independent, initiative-driven teams and translating cutting-edge research into applied platforms. Based in San Francisco, he also authored a comprehensive book on Generative AI, reflecting his focus on both education and real-world impact.
9 years of coding experience
14 years of employment as a software developer
Master of Science - MSc, Computer Science, Master of Science - MSc, Computer Science at Blekinge Tekniska HögskolaDoctor of Philosophy - PhD, Machine Learning, Doctor of Philosophy - PhD, Machine Learning at Wroclaw University of Science and Technology
Github Skills (4)
pytorch10
deep-learning9
deeplearning-ai9
generative-adversarial-network8
Programming languages (3)
Github contributions (5)
jmtomczak/intro_dgm
Jan 2021 - Feb 2022
"Deep Generative Modeling": Introductory Examples
Role in this project:
ML Engineer
ML EngineerContributions:47 commits, 3 PRs, 53 pushes in 1 year 1 month
Contributions summary:Jakub's commits primarily involve implementing and modifying code related to autoregressive models (ARMs), flow-based models, variational autoencoders (VAEs), energy-based models (EBMs), and diffusion-based deep generative models (DDGMs) within the context of deep generative modeling. They demonstrate a focus on the practical implementation of these models, with changes to existing code for improved functionality, and modifications to the RealNVP model, including the addition of components and debugging. The user also implemented an EBM classification loss.
pytorchdeep-generative-modellingdeep-learninggenerative-adversarial-networkgenerative-modeling
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.