Freelancer profile translated to English.

Description

Centralien Engineer and Freelance Data since 2022, I leverage my algorithmic and analytical skills for your technological projects. Depending on your needs, I can work independently on a project from A to Z or as a reinforcement within a technical team.

My positioning allows me to intervene across a wide spectrum:

**- Artificial Intelligence & Deep Learning**: From the R&D phase to the production of complex architectures.

Application areas: Computer Vision (Image Segmentation, Anomaly/Defect Detection), Natural Language Processing (NLP).

**- Data Science, Algorithms & Automation**: Design of tailor-made algorithmic solutions and industrialization of scripts.

Algorithms & Automation: Matching algorithms (database matching), automated generation of technical documents and diagrams, signal processing.

Data & Statistics: Statistical analysis, Feature Engineering, heuristics, data cleaning (Regex), real-time or batch processing.

Technical Stack

**Languages**: Python, Matlab (Notions: R, C++)

**Data Libraries & Ecosystem**: PyTorch, Scikit-Learn, OpenCV, Pandas, NumPy, Matplotlib, Regex

**Environment, Cloud & Tools**: AWS (SageMaker), Jupyter, Git, LaTeX

Languages

French
Native or bilingual
English
Fluent
Portuguese
Conversational
German
Basic

Workplace preferences

Can work on-site

Paris (up to 50km), Marseille (up to 50km), Lyon (up to 10km)

the climate data factory
Machine learning Engineer
ENVIRONMENTAL
July 2025 - January 2026 (6 months)
Paris, France
Deployment of a climate forecasting algorithm on AWS
1. K-mean Clustering
2. Genetic algorithm for Feature selection
3. Forecasting
Machine learning Amazon Web Services AWS SageMaker Python Data science
EssilorLuxottica
Data Scientist - Chaos Theory & Nonlinear Time Series
HEALTH AND WELLNESS
October 2024 - February 2025 (4 months)
Paris, France
Ophthalmic signal analysis

- Advanced signal preprocessing: interpolation, optimal delay estimation using mutual information, attractor reconstruction via phase space diagrams
- Estimation of Lyapunov exponents on different time series using FNN (False Nearest Neighbors) and double regression methods
- Quantification of chaotic behavior using permutation entropy and recurrence plots
RES ELEC
Automation of Electrical Schematics and PDF Reports Generation
CIVIC AND SOCIAL ORGANIZATIONS
April 2024 - June 2024 (2 months)
Paris, France
Creation of a Python algorithm to generate electrical schematics from data, using ReportLab and SchemDraw libraries.
Design of a process to draw and organize electrical schematics, including symbol management and splitting diagrams into multiple frames if necessary.
Automation of PDF report generation including data tables and schematics, with custom layout for each page.

Check out Guilhem's experience

Fabrice

Responsable d’études en traitements d’images - Essilor Instruments

Reviewed on 6/16/2026

Guilhem completed a 5-month mission on a critical component of an image processing algorithm applied to video analysis. He worked autonomously, proposed and successfully tested different technical approaches, and provided a robust solution to the problem. An efficient and committed collaboration until the mission's completion.