Engineering first. AI that ships.
I work at the intersection of data platforms, customer problems and useful software. With 8+ years in Data Engineering, my foundations are Azure, Databricks, Spark/PySpark, data architecture, governance and reliable pipelines. I am building deeper expertise in GenAI, retrieval-augmented generation, evaluation and AI-powered product delivery.
Explore my portfolio & project breakdowns ↗ · Experience · AI Engineering Labs
Some of my best product work is in private repositories. The public case studies describe the problem, architecture, trade-offs and validation approach without exposing private code or confidential data.
| Project | What it demonstrates | Case study |
|---|---|---|
| DossierPro | AI-assisted CV extraction, candidate matching, asynchronous jobs and product workflows | View case study |
| Retail Churn Copilot | Databricks Apps, AI/BI Genie, SQL warehouses and user-facing analytics | View case study |
| SA-OS | Customer discovery, architecture challenge, evidence and reusable AI-assisted SA workflows | View case study |
| Mission Radar | API integration, opportunity tracking, enrichment and operational security | View case study |
| BLOOM Terminal | Cloud/data source monitoring, project watchlists and research workflows | View case study |
| PokéValue | Vision/OCR, entity resolution, multi-source pricing and mobile-first product engineering | View case study |
- COVID-19 Data API — historical Python ingestion + API project, now archived.
- World Inequality Visualizations — D3.js, visual analytics and data storytelling.
- Hex Game AI — Python algorithms, AI agent and graphical client.
- Brent Oil Forecasting — historical time-series Prophet notebook.
Historical projects are presented as learning evidence, not as recently productionised applications.
| Lab | Goal | Evidence to publish |
|---|---|---|
| RAG Evaluation Studio | Retrieval, answer grounding and source citations | recall@k, citation faithfulness, unknown-question behavior, cost and latency |
| Production ML Pipeline | End-to-end reproducible training and serving | baseline, data leakage checks, MLflow, drift monitoring |
| Forward Deployed AI Workflow | A customer-facing workflow built from a realistic business brief | architecture, safe tools, tests, runbook and working demo |
These are planned projects, not yet completed public repositories. See the detailed roadmap.
Python · PySpark · Spark SQL · Databricks · Azure · Delta Lake · Data Governance · CI/CD · TypeScript · RAG · GenAI Evaluation · Cloud Architecture
My approach: understand the user → design the data/AI system → test realistic failure modes → ship a usable product → measure and improve.
Based in France. Interested in applied AI, data platform engineering, technical customer-facing and Forward Deployed Engineering opportunities.