cv
Contact Information
| Name | Louis Grenioux |
| Professional Title | Research Fellow |
| lgrenioux@flatironinstitute.org |
Professional Summary
Machine learning researcher working at the intersection of deep probabilistic modelling and statistical inference, with a focus on Markov Chain Monte Carlo methods enhanced by deep generative models.
Experience
-
2026 - present New York, USA
Research Fellow
Flatiron Institute
- Using ML-enhanced sampling techniques to solve scientific problems and improve Bayesian inference
-
2022 - 2022 Palaiseau, France
Research Assistant
CMAP, École Polytechnique
- Worked under the supervision of Marylou Gabrié and Sylvain Le Corff
- Developed Markov Chain Monte Carlo algorithms enhanced with deep learning
- Designed new algorithms for energy-based model estimation using normalizing flows
-
2021 - 2023 Paris, France
Full Stack Data Scientist
Freelance
- Developed a web platform to use AI tools
- Scaled proofs of concept for production usage
-
2020 - 2020 Paris, France
Research Engineer Intern
Linagora
- Created AI tools for legal audit automation
- Developed a data pipeline from scratch (data engineering, modelling, data visualisation, model evaluation)
- Matched state-of-the-art performance
-
2014 - 2015 Gurgaon, India
Software Engineer (Part-time)
YU Televentures
- Developed the Android 6.0 production OS
- Developed open-source tools for the community
Education
-
2022 - 2025 Palaiseau, France
PhD
CMAP, École Polytechnique
Deep Learning and Statistics
- Supervised by Marylou Gabrié and Éric Moulines
- Research at the crossroads of deep probabilistic modelling and statistical inference through Markov Chain Monte Carlo
- Created new sampling algorithms leveraging deep generative models such as normalizing flows and diffusion models
- Visited José Miguel Hernández-Lobato at the University of Cambridge for 4 months
- Co-organised the 2nd edition of the Frontiers in Probabilistic Inference workshop at NeurIPS 2025
- Publications: 1 ICLR paper, 2 ICML papers (including 1 spotlight), 1 ICML workshop paper, 1 ICLR workshop paper
-
2021 - 2022 Palaiseau, France
MSc
École Polytechnique (IP Paris)
Data Science
- Master Year 2 Data Science of Institut Polytechnique de Paris
- Coursework: Deep Learning, Reinforcement Learning, Computer Vision, Natural Language Processing, Generative Models, MCMC, Optimal Transport, Optimization
-
2019 - 2022 Palaiseau/Évry, France
MSc (Grande École engineering master)
Télécom SudParis
Statistics and Computer Science
- Valedictorian of the promotion TSP2022
- Coursework: Statistics, Data Analysis, Data Mining, Bayesian Inference, Stochastic Processes
- Programming: Java, C, Unix, Web Frontend/Backend, SQL
Awards
-
2022 Hi!ckathon March 2022 — 1st Innovation, 1st Scientific Approach, 2nd Business Opportunity
Hi!Paris (Team Hi!Gency)
2nd edition of Hi!Paris’s high-level hackathon on building AI for climate using computer vision.
-
2022 Generative Modelling for Financial Losses — 1st Place
BNP Paribas and École Polytechnique
1st position out of 60 teams in an AI competition on improving financial stress-test methods using Generative Adversarial Networks.
-
2021 Cassiopée 2021, AI towards Trading — 1st Place
Télécom SudParis
Project leader. Built a real-time trading bot using sequential Monte Carlo (without neural networks). 1st position out of 88 teams. Code.
Projects
-
AI4IA — Artificial Intelligence for Industrial Application
AI competition organised by SKF and hosted by Agorize. Code.
-
Hi!ckathon March 2021 — Team CorAI
1st edition of Hi!Paris’s high-level hackathon. Building AI for energy efficiency using reinforcement learning. Code.
-
AutoJudge
Machine learning model predicting the outcome of US Supreme Court judgments. Code.
-
ArXiv Graphs
Visualising arXiv authors with graphs. Code.
-
Xiaomi SDM439
Open-source development for Xiaomi products using the SDM439 chipset (Redmi 8, 8A, 8A Pro, 8A Dual, 7, 7A). Code.
-
Tinno MSM8937
Reference code for open-source development on Tinno’s ODM MSM8937 devices, including YU Yureka Black (garlic), Wiko Wim Lite (wimlite), and Micromax HS2 (hs2) / HS3 (hs3).
-
AOSParadox
First CAF-based ROM directly mirrored on CodeAurora. Winner of the YU-AOSP-M challenge on 3 devices out of 3. Core developer alongside co-founder Rutash Joshipura. Supported devices include OnePlus One (bacon), OnePlus X (onyx), and various YU devices (Yureka, Yureka Plus, Yureka S, Yureka Black, Yuphoria, Yunique). Code.