Freelancer profile translated to English.
Back to original languageAbout Théo
Computer Vision Engineer | Ex-UC Berkeley Research | OCR (DGFiP, Tessi) | Medical Tech
You are looking toautomatically analyze, classify, or segmentyour complex images?
A graduate ofEPITA(AI & Computer Vision) and having passed through the University ofBerkeley(San Francisco), I help companies in thedesign, training, and deployment of robust AI models.
My expertise:
- **Dataset**: collection (scraping), preparation, and structuring (Python, S3, PostgreSQL).
- **Training**: development of ML / DL models (classification, segmentation, detection, OCR on documents)
- **Production & API**: integration via FastAPI, Cloud deployment (AWS, Docker)
- **Optimization & Scalability**: fast, robust models adapted to your production needs
Main Technologies:
Python / PyTorch / Scikit-learn / OpenCV / FastAPI / Docker / AWS / Numpy / Pandas
Why work with me?
- Solid training:EPITA + Berkeley(experience in a demanding and international scientific environment)
- End-to-end vision: from raw data to production model.
- Concrete results: reliable models usable in real conditions.
- **Pedagogy and clarity**: I explain each step simply, even for non-tech people.
📩 **Let's talk about it**: let's turn your ideas into high-performance, deployable AI models.
French
Native or bilingual
English
Fluent
Remote only
Primarily works remotely
Experience
- ONEEXComputer Vision Engineer | Sovereign Identity VerificationTECHJanuary 2026 - Today (5 months)Paris, FranceDevelopment of a high-performance OCR solution for processing identity documents on CPU in 1s.
- Performance Optimization: Redesign of the OCR architecture, reducing inference time to 1s (from ~10s previously) on CPU via ONNX and OpenVINO.
- Edge AI & Sovereignty: Design of an entirely local processing pipeline (on-device/no-cloud), ensuring the total security of multispectral data (VIS, IR, UV).
- Industrial Scaling: Automation of classification and extraction for over 200 document types across 50 countries with real-time homographic alignment.
- FemnovComputer Vision for Real-Time Ultrasound | EndometriosisMEDICALNovember 2025 - December 2025 (1 month)Paris, FranceDevelopment of a real-time AI solution for analyzing ultrasound videos, dedicated to assisting in the diagnosis of endometriosis. Design of lesion detection and segmentation models on video streams with integrated tracking, and deployment in a containerized production API for clinical use.
- TessiExpert OCR Audit | Tessi (European Leader)DIGITAL AND ITOctober 2025 - November 2025 (1 month)Lyon, FranceConsulting with the technical management of Tessi, a European leader in document processing, on the evolution of their OCR engine.Audit of the existing architecture (preprocessing, OCR models, postprocessing), benchmarking against the state-of-the-art, and recommendations for the next generation of the platform.
Reviews
M
Marilyn
CEO Femnov - Femnov
Reviewed on 2/9/2026
Théo assisted us with the technical scoping and architecture of our medical AI solution. His expertise in Machine Learning / Deep Learning (Computer Vision) is solid and his recommendations are always relevant and actionable.
Very proactive, he anticipates friction points and challenges our choices with pedagogy.
Smooth communication and impeccable collaboration with our CTO.
I recommend him without hesitation. And thank you again Théo for this work!
Charles André
WearIT
Reviewed on 5/17/2025
Surprised by the quality of this freelancer!
I wasn't expecting much, but I was very surprised by the quality of the work provided by Théo. Quick responses, proactivity, advice to improve my project... His work has considerably improved the quality of my AI pipeline and generated a significant return on investment.
I highly recommend 👍
Recommendations
MR
+1
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
- Computer EngineeringEPITA2024Machine Learning | Deep Learning | Computer Vision