Services
With 15 years of experience in software and AI, I build your applications from idea to production. I focus on delivering practical, high-performance tools that solve real problems.
Example: AI integration into your systems (RAG)
Internal docs → semantic retrieval → context → grounded answers.
- AI Audit / Architecture Package (3–5 days) Diagnostics of AI/LLM pipelines, model strategy, scalable architecture, and ROI recommendations.
- Turnkey AI engine setup RAG, OCR, fine-tuning, evaluation framework, and application integration.
- Prototype delivered in 10 days Business chatbot POC, medical OCR pipeline, automated agent, etc.
Profile
Full Stack Engineer with 15+ years building complex systems across Python and React. Recent hands-on experience with RAG systems, LLMs, and AI-powered platforms at scale. I thrive in fast-paced, loosely-defined environments where I can ship features end-to-end. Passionate about making technical knowledge accessible—I've built document processing pipelines, search systems, and developer-facing tools.
Skills
Portfolio
Two representative, production-focused projects (OCR/LLM, multi-agent orchestration, self-hosted LLM gateway).
IOLSolver
Web application for ophthalmologists to define the parameters of an intraocular lens implant. The calculations require entering dozens of values coming from ocular measurement devices. A multi-agent workflow extracts data from photos of device screens or medical reports, reducing manual entry.
- OCR extraction + LLM interpretation
- Multi-agent orchestration for reliability and coverage
- Big time savings on forms with dozens of fields
OpenGateLLM
API gateway to simplify self-hosted LLM usage (load balancing, cost controls, standardized access).
- Single entry-point across models and servers
- Usage guardrails (quotas, costs, etc.)
- Designed for application integration and production rollout
Experience
AI Platform Engineer
DINUM- Built OpenGateLLM, open-source platform managing multiple GenAI servers.
- RAG systems for context-enriched AI answers using internal documents end-to-end.
- Integrated OCR services for automatic document extraction.
Technical Lead
Flowbird- Led 10-engineer team building mission-critical ticketing for Paris transit.
- Designed Python microservices architecture handling millions of daily transient transactions.
- Technologies: Python, Python, React, Docker, MQTT.
Technical Lead
Hyvilo- Led 10-engineer team building mission-critical ticketing system for Paris transit microservices architecture handling millions of daily transactions.
- Technology Stack: Python, React, Docker, MQTT.
Research Engineer
University- Bioinformatics: built data pipelines for genomic analysis (RNA-Seq, metagenomics).
- Processed large datasets with Python, Pandas, R.
Python Developer
QuarksSafety- Built web scraping and data extraction tools for chemical document analysis.
Co-founder
BlouseBrothers- Built collaborative platform for medical students.
- Full stack development.
Lead Developer
LaboOnline- Document management platform for 500+ medical labs.
- Managed 50+ production servers.
Education
MSc Distributed Systems & Networks
University of Franche-Comté
2004 - 2005
Key Strengths
- RAG & LLM Built production RAG systems at DINUM, integrating document sources for AI-enriched responses
- Full Stack Python backend (Django, Flask, FastAPI) + React/TypeScript frontend across multiple projects
- Document Processing Experience extracting and structuring data from complex technical documents (OCR, parsing)
- Startup Environment Startup experience, comfortable with loosely-defined scope and end-to-end ownership
- Scale Built systems handling millions of transactions; experienced with analytics and monitoring