About Yoann
AI & Fullstack Developer | Python | RAG | Agents
- CMB - Sovereign on-premise RAG on banking corpus. Python, Llama, Qdrant, FastAPI.
- MAX - LangGraph business agent orchestrating email, finance, and calendar. 4 hours saved per day on prospecting.
- Seasonal rental - Airbnb/Booking automation. 99%+ availability, zero manual intervention.
- BotYo - Real-time crypto monitoring pipeline. Automatic Telegram alerts, zero supervision.
- AI / LLM: LangGraph, LangChain, RAG, multi-step agents, Qdrant, Llama, OpenAI, Anthropic
- Backend: Python, FastAPI, asyncio, Node.js, TypeScript, PostgreSQL
- Frontend: React, Next.js
- Automation: REST APIs, webhooks, IT system integrations, event-driven workflows
- DevOps: Docker, Linux, Git, Vercel
- Quantifiable results on every project: -70% processing time, 99%+ automation availability, production-ready banking RAG.
- Production-ready architectures: documented, maintainable, not just POCs that never see the light of day.
- End-to-end: scoping, architecture, code, deployment.
- Experienced with sensitive data and multi-system environments.
French
Native or bilingual
English
Fluent
Experience
- Yostan R&DPython Developer / Event-Driven ArchitectureSOFTWARE PUBLISHINGMay 2026 - Today (1 month)Avignon, France
Real-time market monitoring pipeline connected to Kraken WebSocket
Detection of buy/sell signals on various crypto assets, SQLite historization, FastAPI API, automatic Telegram alerts. Configurable detection logic, basic backtesting, deployment on Linux server. Stack: Python, asyncio, FastAPI, SQLite.Result: automatic market monitoring without manual intervention. - Crédit Mutuel BretagneAI Architect / LLM DeveloperBANKING AND INSURANCEJanuary 2026 - February 2026 (1 month)Vannes, France
On-premise RAG architecture for an internal banking document assistant.
Main constraint:Total data sovereignty, zero transit to an external LLM. Complete pipeline delivered: document ingestion, adaptive chunking, embeddings, Qdrant vector store, hybrid retrieval, prompt orchestration, structured answers via Pydantic. Evaluation loop on reference business questions: retrieval accuracy, answer faithfulness, hallucination detection. Complete technical documentation.Result: MVP operational in production, measurable reduction in internal document search time.Skills: RAG / AI Agent / LangGraph / Qdrant / FastAPI - Yostan R&DAI Developer / Agent ArchitectSOFTWARE PUBLISHINGApril 2026 - Today (2 months)Avignon, France
Design of a multi-agent business agent orchestrated via LangGraph.
Central orchestrator routing to specialized agents: email, finance, calendar, notes. Gmail, Qonto, Dougs, calendar integrations. Telegram conversational interface. Persistent state management, typed tools, run logging, error recovery. Stack: Python, LangGraph, LangChain, asyncio, Pydantic, SQLite.Result: 4 hours saved per day on inbound/outbound prospecting.
Recommendations
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Level I Title. Information Systems Manager, Information Technology Project ManagementCentre d'Etudes Supérieures Industrielles2016Titre de niveau I . Manager des systemes d'informations, Gestion de projets en technologies de l''information
- Level II Title. Software Engineering Manager, Information Technology Project ManagementCentre d'Etudes Supérieures Industrielles2014Titre de niveau II Responsable en ingénierie des logiciels, Gestion de projets en technologies de l'information