My journey started with economics, but curiosity led me to explore the technical side of things. I started coding during COVID lockdown and never stopped since then. This exploration has shaped my career path in unexpected ways, combining my analytical background with software engineering.

Currently: AI/ML Engineer at Iceberg Data Lab, working on LLM-based solutions for ESG data extraction and analysis.

Languages & Frameworks:

• Python (primary) - FastAPI, Django, LangChain, Scikit-learn, Pandas, NumPy
• TypeScript & JavaScript - React 19, Next.js, Node.js, Bun
• Rust/Go (occasional projects)

AI & ML:

• Large Language Models (LLMs)
• RAG (Retrieval-Augmented Generation) systems
• Vector databases - Weaviate
• AI SDK, LangChain, prompt engineering

DevOps & Infrastructure:

• Docker & Kubernetes - production deployments
• ArgoCD - GitOps workflows
• Apache Airflow - workflow orchestration
• CI/CD pipelines
• Monitoring & observability

Databases & Storage:

• PostgreSQL, Elasticsearch, Redis
• S3 object storage
• Vector databases (Weaviate)

A full-stack AI chat application with advanced document processing capabilities:

• RAG-powered document chat - Upload PDFs and query them using semantic search
• Multi-agent analysis - Hierarchical agent system for comprehensive document analysis
• Batch questioning - Ask multiple questions across documents simultaneously
• Tech: Bun, React 19, AI SDK, Weaviate, SQLite, S3 storage
• Export conversations to PDF/DOCX/Markdown/JSON

Transform markdown notes into Anki flashcard decks automatically. Write flashcards in your notes using YAML blocks and generate .apkg files ready for import.

• Tech: Python, Anki
• Personal knowledge management workflow

Small library to easily interact with vector stores (Weaviate) and perform RAG operations.

• Tech: Python, Langchain, Weaviate

My development environment configuration, managed with GNU Stow.

• Tech: Lua (Neovim config), shell scripts

Neovim plugin for LLM integration directly in the editor.

• Tech: Lua, Neovim

AI/ML Engineer @ Iceberg Data Lab (Nov 2022 - Present)

• Designed and deployed LLM + RAG solution for automated ESG data extraction from semi-structured reports
• Built FastAPI services for structured data exposure
• Managed Kubernetes deployment with ArgoCD, Docker orchestration
• Deployed Apache Airflow for job orchestration
• Implemented monitoring and continuous evaluation systems for production workloads

Previous: Economics background - Paris School of Economics, Paris 1 Panthéon-Sorbonne (Econometrics & Statistics)

• Large Language Models & generative AI applications
• RAG systems and semantic search
• MLOps & LLMOps best practices
• Cloud-native architectures (Kubernetes, containerization)
• Developer tooling & productivity
• Economic data analysis & visualization

• Blog: https://blog.romainjouhameau.com/
• LinkedIn: Romain Jouhameau
• Email: Available on my blog or CV