Dorian Desblancs is a Senior Research Engineer based in Paris with seven years of experience at the intersection of AI/ML, signal processing, and software engineering. He has advanced speech and music AI from research to production across teams at Deezer and Meta (Voice AI / FAIR), including work on self-supervised learning and novel audio features. Dorian contributes to notable open-source projects—having improved testing and backend compatibility for Deezer’s widely used Spleeter source-separation library—bringing rigor to engineering and test automation. Fluent in both French and English and trained in computer science (McGill) and applied mathematics (ENS Paris-Saclay), he blends theoretical depth with pragmatic engineering. Colleagues describe him simply as a "hacker"—a hint at his hands-on, systems-level approach to building robust ML-driven audio systems.
Deezer source separation library including pretrained models.
Role in this project:
Back-end Developer & Test Automation Engineer
Contributions:8 commits, 8 PRs, 2 pushes in 10 months
Contributions summary:Dorian primarily focused on refactoring and adapting the existing testing framework to accommodate changes in the project's backend. They removed a specific backend (Librosa), indicating efforts to streamline dependencies. Furthermore, they modified function return types and other files to align with the backend changes, ensuring tests remain relevant. The user also addressed code formatting inconsistencies, contributing to code quality through black and isort.
Code for reproducing experiments and figures of the Zero Note Samba paper.
Contributions:62 commits, 2 PRs, 52 pushes in 3 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.
Request Free Trial
Dorian Desblancs - Senior Research Engineer at Meta