Guillaume Dumont
AI Software Developer at CIMMI
Quebec, Canada
Summary
🤩
Rockstar🎓
Top School
Top expert inDeep Learning and Computer Vision Technologies
Guillaume Dumont is an AI software developer with 11 years of engineering experience bridging physics, computer vision, and machine learning, currently focused on production-grade AI at CIMMI in Quebec. A former physics MSc, he brings rigorous analytical thinking to designing efficient software and data-driven solutions across lidar, vision, and simulation domains. His open-source contributions to high-profile projects like Caffe, Caffe2 and Microsoft vcpkg show deep expertise in build systems, cross-platform compatibility and ML ops—particularly improving Windows support and packaging for complex dependencies. Comfortable leading implementations and mentoring teams, he thrives on challenging problems and delivering robust, maintainable systems under real-world constraints. He stays current with emerging tools and often operates at the intersection of research prototypes and deployable software.
11 years of coding experience
11 years of employment as a software developer
Bachelor of Science (B.Sc.), Physics, Bachelor of Science (B.Sc.), Physics at Université LavalMaster of Science (M.Sc.), Physics, Master of Science (M.Sc.), Physics at Université de MontréalPolyvalente L'Ancienne-Lorette
French, English
Github Skills (33)
caffe10
c-language10
lib10
pytest10
python10
windows10
machine-learning10
cmake10
deep-learning10
build-automation10
ccl10
vcpkg10
package-manager10
cprogramming-language10
test-automation10
Programming languages (11)
Github contributions (5)
BVLC/caffe
Oct 2016 - Aug 2017
Caffe: a fast open framework for deep learning.
Role in this project:
ML Engineer & DevOps Engineer
ML Engineer & DevOps EngineerContributions:149 commits, 77 PRs, 41 pushes in 10 months
Contributions summary:Guillaume contributed to the project by fixing testing issues related to temporary file handling on Windows. They also worked on enhancing Python 2/3 compatibility for utility scripts, and adding support for Python 3 and NCCL within the project's core functionality, specifically related to performance improvements. Furthermore, the user updated the boost dependencies and made adjustments to ensure explicit conversion of std::string to boost::python::object.
pytorchvisiondeep-learningmachine-learningcaffe
facebookarchive/caffe2
Jul 2017 - Feb 2018
Caffe2 is a lightweight, modular, and scalable deep learning framework.
Role in this project:MLOps Engineer & Build Automation Engineer
MLOps Engineer & Build Automation EngineerContributions:14 commits, 10 PRs, 20 comments in 7 months
Contributions summary:Guillaume contributed to the Caffe2 framework by addressing build system issues, particularly on Windows platforms. They added support for the Ninja generator, fixed issues related to finding and integrating external libraries (glog, protobuf, LMDB, MKL) in the build process. They also improved the project's compatibility with the Windows environment, including addressing MKL and glog-related problems. The contributions focused on build and dependency management and improving cross-platform compatibility.
pytorchscalablecaffe2deep-learningml
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.