Freelancer profile translated to English.
Back to original languageAbout Youness
I am Youness, a data scientist with over 8 years of experience in the field. My background is marked by strong expertise in Natural Language Processing (NLP), Machine Learning, and Computer Vision. I have had the opportunity to work on challenging projects in various companies such as Adeo, Trustpair, and 2OS/Fortia Financial Solutions, where I applied my skills to diverse problems.
At Adeo, I led the automation of product referencing using Large Language Models (LLMs) and Generative Artificial Intelligence. I also set up a data annotation platform and contributed to building a product-to-usage linking system.
At Trustpair, my role involved automating the supplier verification process by developing information classification and matching systems.
At 2OS/Fortia Financial Solutions, I worked on creating information extraction pipelines from financial documents and improved the performance of a sentence classification system using convolutional neural networks. I also developed a "no-code" platform to facilitate access to Natural Language Processing tools.
My academic background, with a Master's degree in Mathematics, Vision, and Learning from ENS Paris-Saclay and a Master of Science from Ecole Centrale Paris, provided me with a solid foundation for my career.
French
Native or bilingual
English
Native or bilingual
Can work on-site
Paris (up to 50km)
Experience
- BUREAU VERITASMLOps Engineer - Computer Vision & Drone ImageryTRANSPORTATIONOctober 2025 - Today (8 months)Courbevoie, FranceAs part of the industrialization of ML solutions for drone inspection, I am responsible for the MLOps architecture and the optimization of video processing workflows.Responsibilities and Achievements:
- Training Pipeline Orchestration: Design and implementation of automated pipelines for training and retraining computer vision models, ensuring rapid iteration on detection algorithms.
Model Lifecycle Management:- Setup of a Model Registry to centralize and version models.
- Implementation of metric and hyperparameter logging to ensure complete traceability and reproducibility of experiments.
Inference Optimization (Batch Processing):- Development of batch inference architectures for massive video data processing.
- Key Impact: Drastic reduction in processing time for large assets, from 24 hours to only 6 hours (75% performance improvement).
- KPMGGenerative AI EngineerCONSULTING AND AUDITSDecember 2024 - September 2025 (9 months)Courbevoie, France
- Creation and deployment of several RAG (Retrieval Augmented Generation) agents that have significantly improved information retrieval and generation for legal and audit professionals.
- Development and implementation of robust evaluation pipelines using "LLM-as-a-judge" metrics to measure the accuracy and relevance of AI-generated content.
- Building sophisticated multimodal indexing pipelines for vast document sets, enabling efficient processing and retrieval of information from tables, figures, and visual elements.
- Utilization of Azure Search, Celery, and OpenAI models to build scalable and performant AI systems.
- ADEOLead Data ScientistE-COMMERCEOctober 2022 - October 2024 (2 years and 1 month)Lille, France
- Development of a product feature extraction pipeline using state-of-the-art language models (LLMs, including Google's Gemini), few-shot prompting techniques, and self-verification. This resulted in a 19% increase in completeness and a 16% increase in views for enriched products. The solution was deployed in five countries (France, Spain, Portugal, Italy, Poland).
- Creation of a textual product classifier that automatically maps product information to the Adeo DIY taxonomy. This classifier is the result of a weekly automated training pipeline using product catalog data. It was deployed as a real-time API on GCP's Vertex MLOps platform.
- Collaboration with a cross-functional team to create an online data labeling tool. Several data annotation campaigns were coordinated with the help of business experts from multiple European business units.
- Implementation of a distributed logging, tracing, and APM solution using Datadog, integrated with ServiceNow, for Machine Learning applications. This enabled early detection and rapid resolution of production incidents, ensuring SLA compliance and maintaining high user satisfaction.
- Mentoring and coaching several data scientists and machine learning engineers.
Recommendations
+3
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Master's degree, MVA (Mathematics, Vision, Learning)ENS Paris-Saclay2016Master's degree, MVA (Mathematics, Vision, Learning)
- Master of ScienceEcole Centrale2016Master of Science in Applied Mathematics