Corentin Barreau
Software Engineer at Mistral AI
Sarlat-la-Canéda, Nouvelle-Aquitaine, France
Summary
👤
Senior🎓
Top SchoolCorentin Barreau is a pragmatic software engineer with 11 years building large-scale web archiving, crawling and edge AI systems, currently contributing to Mistral AI after leading crawling engineering at the Internet Archive. He combines deep hands-on experience writing and optimizing distributed browser-based crawlers (brozzler) and mass scraping pipelines that processed billions of URLs with embedded ML work integrating NCNN into DeepDetect for efficient edge inference. A digital archivist and open-source enthusiast, he has repeatedly rebuilt critical ingestion pipelines (e.g., INA’s Twitter archiver) and produced datasets and tooling for AI training and live detection on constrained hardware. Based in Nouvelle-Aquitaine, France, he blends low-level performance tuning, full-stack engineering and data acquisition expertise—often surfacing subtle DOM interaction and visibility fixes that improve real-world crawling reliability.
11 years of coding experience
1 year of employment as a software developer
EpitechBaccalauréat Sciences de l'Ingenieur, Baccalauréat Sciences de l'Ingenieur at Lycée Pré de Cordy
Github Skills (23)
pytorch10
javascript10
c-language10
web-crawler10
webscraper10
tensorrt10
machine-learning10
webscraping10
browser-automation10
crawling10
ncnn10
deep-learning10
neural10
neural-network10
crawler10
Programming languages (12)
Github contributions (5)
jolibrain/deepdetect
Sep 2018 - Dec 2018
Deep Learning API and Server in C++14 support for PyTorch,TensorRT, Dlib, NCNN, Tensorflow, XGBoost and TSNE
Role in this project:
ML Engineer
ML EngineerContributions:46 commits, 9 PRs, 9 comments in 3 months
Contributions summary:Corentin primarily contributed to adding and improving the NCNN backend support within the DeepDetect project. They integrated NCNN, a high-performance neural network inference engine, by adding new files, modifying existing ones to include USE_HDF5 flags, and developing unit tests. These changes enable the project to utilize NCNN for model inference, demonstrating expertise in deep learning model deployment and optimization.
xgboostncnnimage-segmentationneural-netsvisual-recognition
internetarchive/brozzler
Sep 2019 - Nov 2019
brozzler - distributed browser-based web crawler
Role in this project:Full-stack Developer
Full-stack DeveloperContributions:14 commits, 9 PRs, 12 comments in 2 months
Contributions summary:Corentin primarily focused on enhancing the functionality of the Brozzler crawler's behavior, specifically within the context of web page interactions. The commits demonstrate the user's work on adding support for specific web page elements, fixing bugs related to closing selectors, and optimizing element visibility checks. They also made contributions to improve the crawler's interval speeds and overall behavior.
browserweb-crawlerweb-basedcrawlingcrawler
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.