Freelancer profile translated to English.

Description

Data scientist for 5 years after a PhD in numerical physics, I have acquired expertise in the design of machine learning or deep learning systems. Over the years, I have explored many types of data (textual, geographical, time series) in many environments (GCP, dbt, dataiku), making me a versatile data science professional from POC to industrialization.

I am available to assist you with your data projects as a Data Scientist or ML Engineer. I master the main Python libraries as well as cloud environments like GCP or Dataiku.

Languages

French
Native or bilingual
English
Fluent

Workplace preferences

Can work on-site

Paris (up to 50km)

bnpparibas
Senior Data Scientist/MLE
BANKING AND INSURANCE
November 2023 - May 2025 (1 year and 6 months)
Paris, France
Design and deployment of a Retrieval-Augmented Generation (RAG) system for information extraction from unstructured documents, from ideation to production:
- LLM Infrastructure: setup and administration of an internal server dedicated to language model inference (vLLM).
- Information Retrieval: combined implementation of vector and lexical search to accurately identify relevant passages.
- Data Structuring: design of optimized prompts to generate structured data from free text.
- Automated Evaluation: development of specialized agents to test and validate the performance of each component of the algorithm.
- Results Persistence: integration of outputs into a database (OracleDB) via SQLAlchemy and Alembic.
- Interoperability: exposure of a REST API (FastAPI) for integration with other internal BNPP CIB solutions.
- User Experimentation: creation of interactive fronts (Streamlit) to test and refine the system.
- Industrialization: full integration into the CI/CD pipeline (Docker, Jenkins, Kubernetes).
MLOps RAG SQLAlchemy LLM ChromaDB
INSEP
Data Scientist
SPORTS
November 2022 - November 2023 (11 months)
Paris, France
Analysis of training data for female athletes to improve training personalization based on hormonal cycles.
- Collection of health information through daily questionnaires
- Analysis of cycling training data (heart rate and power sensors)
- Design of a training load measurement using Hidden Markov Models
- Presentation of results at an international conference (ICSS 2023) and writing a scientific article.
Time Series Pandas Python Data science Scientific article writing
Axionable
Senior Data Scientist
DIGITAL AND IT
January 2019 - June 2021 (2 years and 5 months)
Paris, France
Carried out numerous projects including:
- Predictive maintenance system based on maintenance reports
- Synthesis of stories for children using deep learning
- Classification of sustainable patents
- Prediction of trends in the food industry
NLP Deep Learning Dataiku MLOps Machine learning