Guillaume Salha-galvan
Machine Learning Engineer at SJTU
Shanghai, Shanghai, China
Summary
👤
Senior🎓
Top SchoolGuillaume Salha-galvan is a Machine Learning Engineer and Research Scientist at Deezer in Paris, combining applied product-facing data science with Ph.D.-level research at École Polytechnique. He specializes in deep learning, graph representation learning and recommender systems, with recent work improving GNN-based models for music information retrieval and recommendation. Guillaume has an 11-year multicultural engineering background including roles in Canada, China, and long-standing contributions to the popular PyTorch Geometric library—notably on variational graph autoencoders and related tooling. He holds top-ranked degrees from École Normale Supérieure Paris-Saclay and ENSAE Paris in mathematics, machine learning and data science. Comfortable bridging research and production, he routinely collaborates with product, design and engineering teams to deliver measurable product impact.
11 years of coding experience
Master of Engineering (M.Eng.) in Data Science (Track: Statistics and Machine Learning), Master of Engineering (M.Eng.) in Data Science (Track: Statistics and Machine Learning) at ENSAE ParisDoctor of Philosophy (Ph.D.) in Computer Science / Machine Learning, Doctor of Philosophy (Ph.D.) in Computer Science / Machine Learning at École PolytechniqueMaster's degree (M.Sc.) in Mathematics, Computer Vision and Machine Learning (M.V.A.), Master's degree (M.Sc.) in Mathematics, Computer Vision and Machine Learning (M.V.A.) at École Normale Supérieure Paris-Saclay
Github Skills (8)
geometric-deep-learning10
pytorch10
machine-learning10
deeplearning-ai10
deep-learning10
graph-neural-network10
python10
autoencoder10
Programming languages (2)
Github contributions (5)
pyg-team/pytorch_geometric
Jun 2020 - Sep 2020
Graph Neural Network Library for PyTorch
Role in this project:
ML Engineer
ML EngineerContributions:2 reviews, 7 commits, 2 PRs in 2 months
Contributions summary:Guillaume primarily contributed to the `pytorch_geometric` repository by modifying autoencoder-related files and example scripts. The changes involve the implementation and modification of variational autoencoders (VGAE) and related components within the graph neural network library, including changes to the encoder, KL loss calculation, and reparameterization. The user's updates also include modifications to the test suite, and overall indicate a focus on model development and optimization within the field of geometric deep learning.
pytorchgraph-convolutional-networksgeometric-deep-learningdeep-learningneural-graph
deezer/deezer.github.io
Jun 2020 - Jan 2023
Research team website
Contributions:64 commits, 43 PRs, 62 pushes in 2 years 7 months
jekyll-sitedeezer
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.