France
Cofounded an AI & Web3 services company, generating €74.9k turnover in 12 months (€34.9k personally). Key Projects: ● Internal RAG Platform for Greenty - Built a document-based AI assistant supporting 100+ concurrent users, reducing manual consulting workload. Stack: LangChain for LLM orchestration, Qdrant vector database, OpenAI embeddings, custom OCR models (MiniCPM-V, Qwen-VL) for invoice data extraction. Backend: FastAPI, PostgreSQL, Redis. ● Automated DeFi Arbitrage System - Developed a live arbitrage bot generating 35% annual ROI on a $900k market-neutral portfolio. Automated low-frequency opportunities across DeFi protocols using Node.js, Solidity, and ethers.js. ● E-commerce Chatbot for LEMAITRE Sécurité - Built a conversational AI assistant for a 400+ product catalog with hybrid RAG architecture (semantic search + exact filtering). Stack: LangChain, FAISS vector search, Cohere Rerank, Sentence Transformers for embeddings. ● RAG-as-a-Service Platform (R&D) - Developed a multi-tenant SaaS platform enabling any company to deploy AI assistants with retrieval on both internal documents and product catalogs. Built with a cost-effective and scalable stack: LangChain, Qdrant vector database, FAISS, Sentence Transformers, FastAPI async backend, Next.js frontend. Integrated 350+ e-commerce data sources via Airbyte. Architecture: Docker microservices, WebSocket streaming, complete multi-tenant isolation.
Member of the strategic orientation committee, contributing to the long-term vision and strategic decisions of Nsigma Junior-Entreprise.
Led a student-run consulting firm of 34 members, ranked top 30 Junior-Entreprises in France and ISO 9001 certified. ● Managed a €100k annual budget and oversaw 10+ client projects across various industries ● Represented the organization at national conferences and handled commercial negotiations with clients ● Coordinated strategic development and ensured quality standards compliance
Developed a web application for Michelin, deployed at a professional trade show and used live by visitors during the event. Built with React, Node.js, and REST API integration.
Deployed the internal RAG platform initially developed at Ailog into production. The platform automated manual consulting tasks and demonstrated AI capabilities that established recurring revenue models for the company. ● Implemented advanced OCR on complex documents for automated invoice processing using MiniCPM-V and Qwen-VL models ● Built scalable infrastructure supporting 100+ concurrent users with FastAPI backend, PostgreSQL, and Redis caching ● Reduced manual workload through document processing and retrieval with LangChain and Qdrant vector database
Developed quantitative models and trading strategies in the decentralized finance space. ● Built quantitative models analyzing macroeconomic factors and their impact on digital asset markets ● Conducted on-chain data analysis to identify market inefficiencies and trading opportunities ● Developed smart contracts in Solidity and interacted with DeFi protocols using Node.js for strategy implementation