Freelancer profile translated to English.

Description

🎓 Data Scientist with over 6 years of experience, specializing in Timeseries and Machine Learning, I support my clients in the design and deployment of AI solutions adapted to their needs.

⛏️ My expertise:

**Machine Learning & Computer Vision**: Regression, Forecasting, Timeseries, anomaly detection, image classification, segmentation, CNN, U-Net (XGBoost, Scikit-learn, Tensorflow, Keras, CUDA...)
**Data Engineering & API**: ETL Pipelines, SQL/MySQL, BigQuery, FastAPI
**Cloud & MLOps**: Deployment of models on cloud instances, CI/CD management with GitHub Actions, Docker, model monitoring and versioning via MLflow or ZenML.
**Data Visualization**: Interactive and dynamic dashboards (Power BI, React + Shadcn, Plotly)

💪 My work methodology:

**Briefing and needs analysis**: understanding the context, scoping, feasibility study.
Research and proof of concept– exploration of the state of the art, proposal of approaches and realization of a POC.
**Technical development**: prototyping, API creation, database setup (AWS RDS, Supabase or other).
**Demonstration and validation**: rapid iterations with the client to refine the product.
**Industrialization and deployment**: production deployment on a cloud instance (often AWS).
**Final delivery**: turnkey solution, with an optional React interface for an optimal user experience.

🎁 Bonus: I will create a user-friendly dashboard in React + Shadcn for your project.

🪖 My commitment: to provide you with a reliable, high-performing solution perfectly aligned with your objectives, within the announced deadlines.

👉 Discover my portfolio and request access to the demo of my RAG agent to see my work in real conditions.

Industry field of expertise

Languages

English
Native or bilingual
French
Native or bilingual

Workplace preferences

Can work on-site

Paris (up to 50km)

ENEDIS (Ex ERDF)
Data Scientist
ENERGY AND UTILITIES
September 2022 - Today (3 years and 11 months)
Paris, France
As part of the war in Ukraine and national energy sobriety efforts, participated within Enedis's OPH team in analyzing levers for optimizing electricity consumption across various sectors in France.
• Development and deployment of XGBoost Machine Learning models on MLflow.
• Development and deployment of ML forecasting models to analyze electricity consumption patterns of different consumer segments.
• Design of an ETL pipeline using Airflow, Python, and PostgreSQL:
o Extraction of consumption data from various Hadoop servers.
o Transformation and loading into our Teradata data warehouse.
o Automation of data updates to the Open Data API (FastAPI).
• Creation of interactive dashboards in PowerBI to visualize energy consumption at national, regional, and sectoral levels.
• Writing detailed technical specifications and maintaining data processing packages with Jfrog, Jira.
Python TensorFlow Django API SQL
MediaRGB
Computer Vision Engineer
TECH
February 2022 - August 2022 (7 months)
Paris, France
POC aimed at designing an API capable of transforming user photos into personalized avatars using facial feature classification.
• Training of convolutional neural network (CNN) models withScikitand **Tensorflow**.
• Use of U-Net based semantic segmentation to accurately extract users' hair and facial hair.
• Labeling of thousands of facial images to ensure high-quality training data.
• Image preprocessing with OpenCV.
• Principal Component Analysis (PCA) and KNN classification.
• Deployment of theFastAPIAPI on anAWS EC2instance.
TensorFlow Python OpenCV AWS Computer Vision
Altalia Courtage
Data Scientist
BANKING AND INSURANCE
January 2021 - January 2022 (1 year)
Paris, France
Internal project aimed at replacing an expensive third-party CRM with a proprietary, customizable internal solution.
• Implementation of the client and insurance database withPostgreSQLon AWS RDS.
• Development of the CRM with React and Bootstrap for the front-end using the API solution.
• Complete documentation of API endpoints, database schemas, and system architecture.
• Setup of an autonomous CI/CD pipeline via GitHub and **Docker**.
• Integration of a machine learning pricing model (Scikit-learn/XGBoost) to generate personalized offers based on client profiles.
Python Pandas Gitlab CI/CD Machine learning django