About Thomas
French
Native or bilingual
English
Fluent
Japanese
Conversational
German
Basic
Experience
- Beta.GouvData Engineer | Python | SQL | Analytics EngineerPUBLIC SECTORSeptember 2025 - Today (9 months)Lyon, FranceThe state startup Covoiturage assists local authorities, employers, and decision-makers in promoting short-distance carpooling. Among its services, the startup acts as a trusted third party for collecting and processing carpooling proof.My role is at the intersection of data engineering and data analysis.I have performed the following tasks:- Carpooling data analysis.- Measuring the effectiveness of incentive campaigns.- Reporting to local authorities.
- ÆGIRData Engineer | Python | SQL | Analytics EngineerENVIRONMENTALMarch 2025 - July 2025 (4 months)Lyon, France→ Mission Objective:Automate the collection and analysis of geospatial and meteorological data to optimize flood risk forecasting at the national level.→ Achievements:Designed an automatedPython ETL pipelinecovering all of metropolitan France, with multi-source ingestion (Météo-France API, data.gouv) forreliableandscalableprocessing of environmental data.AdvancedorchestrationviaAirflowand optimized geospatial transformations (land cover, altitudes) usingPostgreSQL/ **PostGIS**, **Duckdb**, and **dbt**, improving data performance and quality.Implemented intelligent scraping (Selenium**, **BeautifulSoup) to enrich datasets, eliminating manual collection time.→ Technical Stack:Python, Airflow, dlt, dbt, PostgreSQL, PostGIS, DuckDB, BeautifulSoup, Selenium, HDBScan
- PyronearData Engineer | Python | SQL | Analytics EngineerENVIRONMENTALSeptember 2024 - Today (1 year and 9 months)Lyon, France→ Mission Objective:Strengthen early wildfire detection throughComputer VisionandDeep Learningmodels optimized by **Generative AI**.→ Achievements:Generation of synthetic data usingStable Diffusionto improve the diversity of use cases and boost the robustness ofDeep Learningmodels.Deployment of a completeData Science/ MLOps pipeline: MLFlow tracking, automated data augmentation, and DreamBooth fine-tuning, enhancing the quality and performance of the final model.Training and calibration ofYOLOandSegment Anything (SAM)models for precise and rapid detection, significantly reducing the false positive rate in **Computer Vision**.→ Technical Stack:Python, PyTorch, MLFlow, YOLO, SAM, Stable Diffusion, HuggingFace, DreamBooth, Computer Vision, Data Scientist
Recommendations
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
- PhD in High Energy PhysicsSorbonne UniversityDoctorat en Physique des Hautes Énergies
Certifications
- PhDSorbonne University2022
- GDPR, AI Ethics and AI ActDataScientest