Freelancer profile translated to English.

Description

I design and deliver intelligent systems in production: LLM agents, AI pipelines, and high-load Python backends.

My approach: combine the rigor of backend engineering (FastAPI, Docker, Kubernetes) with the intelligence of agentic architectures (LLM orchestration, RAG, escalation, human-in-the-loop). I don't do POCs that stay in a notebook — I deliver systems that run.

What I bring concretely:

→ Agents & AI: multi-agent orchestration, ReAct patterns, LLM escalation with HITL, RAG with embeddings, PyTorch, scikit-learn

→ Scalable Backend: Production FastAPI APIs at 15,000 calls/minute across the entire system, worker + message broker architectures, 200,000 processes/day

→ Data platforms: creation from scratch on GCP (BigQuery, Composer, Dataflow), declarative frameworks, ingestion pipelines

→ Tech Lead: functional scoping, team management, SAFe sprints, delivery guarantee

Last mission: Tech Lead Backend at FedEx (30 countries, 15,000 drivers, 600 stations) — route planning system that generated €7 million in savings.

Daily international context (fluent written/spoken English) with teams in India, Germany, UK, Netherlands, and US.

Computer Engineering graduate — Polytech Tours.

Industry field of expertise

Languages

English
Fluent
French
Native or bilingual

Workplace preferences

Can work on-site

Tours (up to 50km), Paris (up to 10km), Le Mans (up to 10km), Orléans (up to 10km)

FedEx
Tech Lead Backend
TRANSPORTATION
February 2024 - Today (2 years and 4 months)
Route planning and tracking system for 15,000 drivers in 30 countries. Backend orchestrating internal expert systems (packages → stops → sequences → FastAPI API). 200k sequences/day, API at 1,000 calls/min. Team management, functional scoping, SAFe sprints. Replacement of the old system — savings of €7 million.
Stack: Python, FastAPI, Pydantic, SQLAlchemy, Redis, PostgreSQL, RabbitMQ, Azure, AKS, Kubernetes.
Python FastAPI RabbitMQ Microsoft Azure Kubernetes
Bforbank
Data Engineer / Tech Lead
BANKING AND INSURANCE
October 2023 - February 2024 (4 months)
Puteaux, France
Creation of the data platform from scratch. Declarative YAML framework (ingestion, dbt processing, outputs), custom Python framework, cloud-agnostic GKE/Kustomize deployment. Detection of social engineering fraud (~€10k/day) via pipeline + alerting.
Stack: Python, dbt, BigQuery, Composer, Dataflow, Pub/Sub, GCS, Looker, GKE, Kubernetes.
Python PubSub Kubernetes BigQuery Google cloud
Cyrès - Solutions Collaboratives, Cloud & Big Data
Data Engineer / ML Engineer
BANKING AND INSURANCE
September 2018 - February 2021 (2 years and 6 months)
Région de Tours, France
Banking fraud detection system using deep learning — 25% reduction in losses. Industrialized microservices architecture on Kubernetes. Stack: PyTorch, scikit-learn, Flask, RabbitMQ, PostgreSQL, Redis, Kubeflow, GCP, Docker.
Kubernetes Pytorch PostgreSQL Redis RabbitMQ