Bart Van Merriënboer
Research Scientist at Google
Montreal, Quebec, Canada
Summary
🤩
Rockstar🎓
Top SchoolBart Van Merriënboer is a research scientist at Google DeepMind with 13 years of experience in deep learning, machine learning systems, and optimization, blending rigorous academic training (PhD with Yoshua Bengio) with production-focused research roles at Google, Meta and Twitter. He has a strong open-source track record contributing to foundational ML tooling like Theano/PyTensor, Blocks, Fuel and PyLearn2, improving core infrastructure such as scan error messages, data pipelines and neural network building blocks. His work spans both algorithmic advances and engineering hardening for real-world workflows, and he applies ML to biodiversity and bioacoustics—an example of translating core research into domain impact. A graduate of a competitive Erasmus Mundus complex systems programme and former TEDxWarwick coordinator, he pairs technical depth with cross-disciplinary leadership and a history of scaling student-organized events.
13 years of coding experience
3 years of employment as a software developer
Bachelor of Science (BSc), Mathematics and Business Studies, First Class Honours, Bachelor of Science (BSc), Mathematics and Business Studies, First Class Honours at University of WarwickBusiness, Israeli politics, Middle Eastern history, Ethics, A (93%), Business, Israeli politics, Middle Eastern history, Ethics, A (93%) at Tel Aviv UniversityDoctor of Philosophy (Ph.D.), Machine learning, A, Doctor of Philosophy (Ph.D.), Machine learning, A at Université de MontréalMaster of Science (MSc), Complex Systems Science (Erasmus Mundus), GPA 3.9/4.0, Master of Science (MSc), Complex Systems Science (Erasmus Mundus), GPA 3.9/4.0 at Ecole polytechniqueMaster of Science (MSc), Complex Systems Science (Erasmus Mundus), A, Master of Science (MSc), Complex Systems Science (Erasmus Mundus), A at University of Gothenburg
Dutch, English, French, Spanish
Github Skills (27)
data-stream10
transformers10
convolutional-neural-networks10
data-pipelines10
python10
machine-learning10
numpy10
lua10
deep-learning10
tensorflow10
neural-network10
data-representation10
data-pipeline10
fuel10
theano10
Programming languages (10)
Github contributions (5)
facebookarchive/bAbI-tasks
Sep 2015 - Jan 2016
Task generation for testing text understanding and reasoning
Role in this project:
Back-end Developer
Back-end DeveloperContributions:16 commits, 1 PR, 9 pushes in 4 months
Contributions summary:Bart primarily contributed to improving the functionality and correctness of the bAbI tasks generation code. Their work includes fixing a randomness bug, adjusting positional reasoning logic, and making the number of decoys configurable for the PathFinding task. The user also addressed several error messages, typos and added a script to check uniqueness of tasks. The changes primarily involved modifications to Lua code related to task generation and configuration.
testingunderstanding
mila-iqia/blocks
Oct 2014 - Mar 2017
A Theano framework for building and training neural networks
Role in this project:ML Engineer
ML EngineerContributions:1 release, 1114 commits, 356 PRs in 2 years 5 months
Contributions summary:Bart made changes primarily related to the Blocks framework for building and training neural networks. They were responsible for implementing and testing the core components of a neural network, including convolutional layers and max pooling, and providing supporting functionalities for various aspects of training. The user also demonstrated knowledge of data handling and integrating the framework with other machine learning components. Their work included several iterations on improving the performance and functionalities for a sequence generator and its components.
pytorchdeep-learningtheanoneural-networksmachine-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.