I industrialize your AI systems
scoping → POC → production
AI & ML Engineer (3.5 years of XP) and 2× AWS Certified, I build and run your AI solutions, on the cloud and on-prem. Concrete, sovereign, high-performing.
Chat with my 3D avatar, check how your needs match my profile, or upload a document for analysis.
For complex or confidential requests, contact me directly.
Talk with my AI avatar, live
Your image here
Your AI POC is ready?
INITIALIZING PIPELINE...
PRODUCTION LIVE
Oops, something went wrong. Try again in a moment.
My Services
From Machine Learning to LLMOps, in the field.
YMUNIT
Technical AI lead for a B2B SaaS platform in GRC and cybersecurity: architecture, standards and roadmap for LLM, RAG and agent projects, from POC to production.
- Sovereign serving of open-weight models with vLLM on H100 GPUs in an air-gapped environment
- 1.8× throughput, -40% inference cost and -45% VRAM footprint
- AWS infrastructure automated with Terraform, Docker and GitLab CI/CD, provisioning cut to under an hour
- Benchmarked around ten models and cut token cost by roughly 30% at equivalent quality
YMUNIT
Designed the AI ecosystem for a document analysis, risk management and cybersecurity platform on a hybrid on-premise/SaaS architecture.
- EBIOS analyses accelerated ×18 and over €250k in annual savings for clients
- Multi-format RAG pipeline with OCR, hybrid search, reranking and citations, Recall@7 of 0.95
- Vulnerability analysis based on CVE and MITRE ATT&CK
- Agentic workflows, tool calling, structured outputs and human validation
SNCF Voyageurs
Predictive analytics applied to railway safety in the context of the Paris 2024 Olympic and Paralympic Games.
- ML models to identify and prioritize at-risk stations
- Streamlit app and real-time monitoring dashboard
- Automated data processing and reporting
CNRS
Predictive modeling for a research project on estimating the lifetime of frozen soap bubbles.
- Structuring experimental data
- Statistical analysis
- Comparing supervised models