Mahmoud Abuzaina
Deep Learning Software Engineer at Intel Corporation
Chandler, Arizona, United States
Summary
🤩
Rockstar🎓
Top SchoolMahmoud Abuzaina is a Deep Learning Software Engineer with nine years of experience building and optimizing ML systems at Intel, based in Chandler, Arizona. He specializes in performance-focused model engineering—contributing to flagship open-source projects like TensorFlow and Intel AI Reference Models by enabling fused Conv and INT8 ops, BF16 support, and adapting RFCN object-detection workflows to TensorFlow 2.0. His background spans computer vision, real-time video processing, and embedded/desktop applications from internships to production-grade deep learning stacks, giving him a rare combination of low-level optimization and applied ML expertise. Mahmoud holds a CS degree from the University of Jordan and an MS from Cal State East Bay, and is known for improving inference efficiency on Intel hardware—work that directly impacts benchmarking and deployment at scale.
9 years of coding experience
2 years of employment as a software developer
Master's Degree, Computer Science, Master's Degree, Computer Science at California State University - East BayBachelor's Degree, Computer Science, Bachelor's Degree, Computer Science at University of Jordan
English, Arabic
Github Skills (15)
neural-network10
object-detection10
machine-learning10
deep-learning10
tensorflow10
edn10
deep-neural-networks10
python9
cprogramming-language9
c-language9
inference9
convolution9
ai8
performance-analysis7
performance-monitor7
Programming languages (4)
Github contributions (5)
intel/ai-reference-models
Jul 2019 - Jan 2022
Intel® AI Reference Models: contains Intel optimizations for running deep learning workloads on Intel® Xeon® Scalable processors and Intel® Data Center GPUs
Role in this project:
ML Engineer
ML EngineerContributions:26 commits in 2 years 6 months
Contributions summary:Mahmoud's commits primarily involve upgrading RFCN scripts to work with TensorFlow 2.0 within the context of object detection. They focused on adapting the RFCN model for benchmark and accuracy testing using TF 2.0 and made updates to related scripts, including accuracy calculations. Additionally, the user made changes to the README, integrating the use of integer scripts and formatting improvements.
optimizationsprocessorstensorflowzoomodel-zoo
tensorflow/tensorflow
Feb 2017 - Oct 2022
An Open Source Machine Learning Framework for Everyone
Role in this project:ML Engineer
ML EngineerContributions:93 reviews, 213 commits, 128 PRs in 5 years 9 months
Contributions summary:Mahmoud's commits focused on enabling and optimizing machine learning operations within the TensorFlow framework. Their primary contribution involved implementing fused operations, specifically enabling and optimizing Conv2D, DepthwiseConv2D, and BiasAdd + FusedHardSwish fusion for FP32/BF16. Further improvements included enabling remaining INT8 ops and oneDNNv3.1 within INT8 Conv operations, demonstrating a focus on performance and efficiency within the framework's capabilities.
pythondata-sciencedeep-learningmlmachine-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.