Experience
Machine Learning Engineer — lambda.ai
Jun 2022 – Oct 2024
- Built and optimized the LLM pipeline for an AI-generated news site (news.lambda.ai)
- Developed GenAI model benchmarks, including an inference benchmark for Stable Diffusion cited by Google Research in the Muse paper
- Published Stable Diffusion finetuning demos on Hugging Face
- Conducted 8 technical interviews, contributing to hiring decisions for ML engineering team growth
Machine Learning Engineer — Deep Voodoo
Jun 2022 – Dec 2023
- Led the transition of a deepfake research pipeline into a scalable, production-ready system deployed in high-profile Hollywood projects
- Improved the robustness and deployment speed of the pipeline by versioning and containerizing model training
- Led R&D projects in image restoration, facial detection for non-human faces and image generation with controllable camera angle
AI Software Engineer — Boston Consulting Group
Jun 2019 – Jul 2021
- Built end-to-end data pipelines using PySpark, Pandas
- Productionized and deployed data applications across diverse client infrastructures, including bare-metal, GCP, AWS, and Azure
- Built a CI/CD pipeline using JFrog to let consultants self-deploy cloud-based coding environments by submitting a Dockerfile via a frontend interface
Data Scientist (1st employee) — Prazely
Oct 2018 – Apr 2019
- Built a content-based recommender system for an e-commerce website using a graph database backend (neo4j)
Research Scientist — Orange Telecom
May 2018 – Sep 2018
- Benchmarked a parameter-free time series classifier against existing models
Freelance Python Developer — Stealth Analytics Startup
Jul 2016 – Sep 2017
- Built a Twitter data analytics web app using Flask
- Implemented unsupervised NLP techniques and graph queries to identify topics and communities of users
Education
M.S. in Mathematics and Computer Science
2017 – 2018 · Université Paris Dauphine - PSL
M.B.A.
2015 – 2016 · Bentley University
M.S. in Management
2011 – 2014 · NEOMA Business School
Articles and Projects
- “All You Need Is One GPU” — Inference Benchmark for Stable Diffusion (2022)
- “Text-to-naruto” — A Stable Diffusion Finetuning Demo (2022)
- “GiftHub” — Collective Decision-Making Tool for Funding Distribution (2021)