Freelancer profile translated to English.

Description

Data Scientist specializing in agentic AI with over 4 years of experience in applied research and the development of distributed intelligent systems. Expert in designing multi-agent pipelines, integrating ML/DL models into orchestrated environments (LangChain, LangGraph), and deploying on modular architectures (FastAPI, MCP servers, Docker containers). Strong command of complex data processing, autonomous decision-making, and LLM+tool workflows, including in constrained environments (Jetson Nano, edge computing). Curious, rigorous, focused on innovation and real-impact systems.

Languages

French
Native or bilingual
English
Native or bilingual

Workplace preferences

Can work on-site

Villeurbanne (up to 50km), Lille (up to 50km), Lyon (up to 50km), Paris (up to 50km)

Pixie-Services
AI & LLMOPS Engineer
DIGITAL AND IT
October 2025 - Today (10 months)
Lille, France
AI and Web Development Engineer
Creation of end-to-end AI solutions
Langchain - Langgraph - NextJS - Graphql Etc...
Next.js GraphQL Langchain Vite.js Github Actions
DECATHLON
AI Engineer
DIGITAL AND IT
October 2025 - Today (10 months)
Lille, France
Development of end-to-end agentic AI systems.
Web Dev
Google Cloud platform
AWS
Langchain Langraph
NextJS
Etc...
Python Langchain LangGraph Google ADK Google Cloud
DAG SYSTEM
AI / LLMOPS ENGINEER
SOFTWARE PUBLISHING
February 2025 - September 2025 (7 months)
Villeurbanne, France
- Design and development from scratch of a distributed, microservices-oriented back-end architecture (FastAPI, GraphQL, MCP servers) for an intelligent tourism platform competing with industry giants.

- Implementation of a multi-agent pipeline (LangChain) for the unification, semantic merging, and contextual enrichment of heterogeneous data (points of interest, events, weather, itineraries…).

- Orchestration of specialized agents (format unification, semantic merging, enrichment via external APIs, dynamic itinerary planning based on weather and preferences).

- Advanced integration of multiple data sources (MongoDB, MySQL, Google Places APIs, MapBox, weather, etc.) and exposure via GraphQL for complex, low-latency queries (<500 ms).
- Contribution to the front-end part (mapping, dynamic filters) to ensure a fluid and interactive user experience.

Result: System capable of centralizing, enriching, and dynamically serving thousands of points of interest, with real-time personalized itinerary generation.
LangGraph Langchain FastAPI Backend Service Development Python Programming