Olivier Dulcy
Paris, Ile-de-France
Summary
🤩
RockstarOlivier Dulcy is a Senior Machine Learning Engineer based in Paris with three years of hands-on experience building and deploying LLM-powered information extraction systems. He has led MLOps and on-prem GPU infrastructure efforts, implemented large-scale models (Llama, SFT/LoRA, QLoRA) and optimized inference with quantization to serve production endpoints on AWS and GCP. As official maintainer of docTR and founder of Tuxae, he blends open-source stewardship with practical product ownership and a penchant for DIY infrastructure—he runs a homelab of Dell servers and has built training platforms on-prem. Olivier’s background spans NLP research, ClearML-driven experimentation, and end-to-end CI/CD for model delivery, making him adept at turning cutting-edge research into reliable, scalable applications.
4 years of coding experience
Github Skills (56)
pytorch10
text-recognition10
ocr10
deep-learning10
text-detection10
optical-character-recognition10
amd9
training-data9
inference9
synthetic-dataset-generation9
data-augmentation9
gpt9
cuda9
transformer8
llama8
Programming languages (3)
Github contributions (5)
odulcy-mindee/doctr
Jul 2022 - Jan 2025
docTR (Document Text Recognition) - a seamless, high-performing & accessible library for OCR-related tasks powered by Deep Learning.
Contributions:3 PRs, 222 pushes, 72 branches in 2 years 6 months
optical-character-recognitiontext-recognitiondeep-learningaccessiblerecognition
odulcy-mindee/augraphy
Dec 2022 - Jan 2023
Augmentation pipeline for rendering synthetic paper printing, faxing, scanning and copy machine processes
Contributions:2 PRs, 2 pushes, 4 branches in 1 month
printingpipelinerenderingscanningsynthetic
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.