Aleks Kamko
AI Infrastructure Engineer at Figma
Santa Monica, California, United States
Summary
🤩
Rockstar🎓
Top SchoolAleks Kamko is an AI infrastructure engineer with 13 years of hands-on experience building production-grade systems and founding engineering teams that ship real products. He’s driven end-to-end ML and systems work—filtering and captioning 650M images, building training infra for 256x H100 clusters, and delivering open-source foundation models that rival Midjourney v6 in aesthetics. Earlier he pioneered low-latency cloud browser streaming at Mighty, hacking Chromium, GPU encoders, and custom networking to achieve tens-of-millisecond latency. Aleks combines research-grade deep learning expertise with pragmatic infra optimization (e.g., >2x PyTorch Trainer/DataLoader speedups and significant cloud cost savings via spot autoscaling). He contributes to ML tooling like BIDMach’s Random Forest improvements, reflecting a long-term interest in efficient, accelerated ML on CPU/GPU. Based in Santa Monica, he pairs UC Berkeley EECS training and teaching experience with a track record of turning exploratory research into high-MRR products.
13 years of coding experience
8 years of employment as a software developer
Master of Engineering Data Science & Systems, Master of Engineering Data Science & Systems at University of California, Berkeley
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (4)
machine-learning10
random-forest10
scala10
data-processing8
Programming languages (15)
Github contributions (5)
BIDData/BIDMach
Sep 2016 - May 2017
CPU and GPU-accelerated Machine Learning Library
Role in this project:
ML Engineer
ML EngineerContributions:77 commits, 13 PRs, 69 pushes in 8 months
Contributions summary:Aleks's contributions focused on improving the Random Forest model within the BIDMach machine learning library. They made various code changes, including making impurity functions static and ensuring classes are serializable. The user also appeared to be involved in the addition of new scripts related to MNIST8M fine-split data processing. These changes suggest a focus on improving the functionality, and usability of the Random Forest model.
cudacpudeep-learninggpuaccelerated
aykamko/parsec-google-cloud-gaming
Mar 2020 - Mar 2020
Contributions:25 commits, 15 pushes in 8 days
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.