Théo Bonzi

Lead Computer Vision Engineer | Deep Learning

Salary / Daily rate

Paris, France

Freelance, open to permanent contract

Can come onsite

Skills

PythonPyTorchMachine LearningAIDeep LearningData EngineeringKerasReact NativePostgreSQLVercelGitRESTful APIsLLMs / Large language modelsReactNumPypandasTensorFlowC#Azurescikit-learnMongoDBAmazon EC2C++DockerKafkaPostmanLinuxSupabaseGitHubSQL databasesCAWS

Languages

French (Native)English

About me

Intro

Computer Vision & Deep Learning expert with 2+ years building production AI systems. Led development of real-time virtual try-on models trained on 3M+ images, achieving 92% OCR accuracy for government digitization projects. Specialized in multimodal AI, real-time inference optimization, and scalable ML pipelines using TensorFlow/Python.

Links

Work experience

Lead Computer Vision & IA

WearIT

Feb 2024 - Present

2 years 9 months

Paris, France

Built from scratch high-performance AI models for virtual clothing try-on. Trained on 3M+ custom-labeled images and optimized for real-time inference (< 2 sec). Currently powering the WearIT mobile experience, redefining how users visualize and interact with fashion online.

AI Research Intern - Astronomy

University of California, Berkeley

Mar 2024 - Aug 2024

6 months

Berkeley, CA, USA

Developed multimodal AI to classify supernovae from massive, real-time astronomical datasets. Collaborated with leading researchers at Berkeley, Caltech, and contributed to the GRANDMA project (NASA/CNRS) for gravitational-wave detection. Presented research results weekly, refining models based on scientific feedback.

Computer Vision Engineer

Direction générale des Finances publiques

Mar 2023 - Mar 2024

1 year 1 month

Paris, France

Built automated OCR pipeline to digitize handwritten tax documents, handling thousands of forms daily. Achieved 92% accuracy, significantly improving administrative efficiency. Successfully navigated technical challenges specific to public-sector needs.

Full Stack Developer Intern

Manger NU!

Sep 2022 - Jan 2023

5 months

Paris, France

Developed internal tools improving sales team workflows and productivity. Collaborated closely with sales staff, continuously refining features based on user feedback. Delivered effective, user-friendly software solutions still in use today.


Education

EPITA: Ecole d'Ingénieurs en Informatique

Diplôme d'ingénieur

2018 - 2024

5 years 11 months

Université du Québec à Chicoutimi

Semestre à l'étranger

2020 - 2020

3 months

Lycée Las Cases

Baccalauréat scientifique

2015 - 2018

2 years 11 months

The best place to find your next move

© 2026. All Rights Reserved.