George Seif
Principal Machine Learning Engineer at Atlassian
Toronto, Ontario, Canada
Summary
🤩
Rockstar🎓
Top SchoolGeorge Seif is a Principal Machine Learning Engineer based in Toronto with 13 years of experience building and shipping production ML and Generative AI products. He has led end-to-end systems from research and model training to scalable deployment and monitoring, most recently architecting AI-powered product initiatives at Atlassian after driving Generative AI efforts and RAG systems at Ada. His background spans AutoML platforms, continuous retraining pipelines for video-scale computer vision, and hands-on performance fixes in open-source projects like a TensorFlow semantic segmentation suite where he addressed training memory leaks. Comfortable bridging technical and non-technical stakeholders, he mixes deep academic training in electrical engineering and machine/deep learning with pragmatic engineering: designing modular LLM chains, automated evaluation tooling, and robust production infrastructures. Colleagues rely on him for technical leadership, system architecture, and translating bleeding-edge models into reliable customer-facing features.
13 years of coding experience
7 years of employment as a software developer
Toronto Metropolitan University
Github Skills (5)
semantic-segmentation10
computer-vision10
tensorflow10
python10
image-processing9
Programming languages (9)
Github contributions (5)
GeorgeSeif/Semantic-Segmentation-Suite
Dec 2017 - Mar 2019
Semantic Segmentation Suite in TensorFlow. Implement, train, and test new Semantic Segmentation models easily!
Role in this project:
ML Engineer
ML EngineerContributions:143 commits, 13 PRs, 117 pushes in 1 year 3 months
Contributions summary:George focused on fixing a memory leak issue within the training process, indicating a direct involvement in model training and optimization. The changes involved modifying the training script to address the memory leak caused by the creation of new tensors during image cropping. This suggests the user was working to improve the training efficiency, likely for semantic segmentation models. The user also added color coding for displaying the results.
trainsemantic-segmentationdeep-learningsemantic-segmentation-modelstensorflow
rgegriff/blokSCAD
Jul 2016 - Sep 2016
Contributions:8 commits in 2 months
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.