Freelancer profile translated to English.

Description

Senior AI Engineer with 10+ years of experience in designing and deploying data-intensive platforms and AI solutions in critical environments.

Expertise in Python / FastAPI / Rust, Cloud (GCP, Kubernetes) and MLOps / LLMOps with a specialization in the industrialization of generative AI solutions (RAG, LangGraph, LLM) in production.

Confirmed technical leadership, with mentoring teams of up to 12 developers and managing complex, high-stakes projects.

Solid track record on large-scale projects: massive migrations (+22M lines, +3.5M contracts) and digital transformation for major organizations (CAC40, public sector).

Key skills:

- AI & LLMs: Design, Integration, evaluation, automation, Fine-tuning,

- Languages & Frameworks: Python (FastAPI), Rust, Langchain, LangGraph, LlamaIndex, Haystack, PyTorch, RAGAS,FastMCP

- Data Engineering (Data pipelines, ingestion, transformation): Pandas, Polars, Airflow, GCP Workflows

- Cloud & MLOps: AWS, GCP, Docker, Kubernetes, MLflow

- Vector DB & Semantic Search: Chroma, FAISS

- Architecture: Distributed systems, APIs, Microservices, Clean, Hexa

Availability: Full remote only

Industry field of expertise

Languages

French
Native or bilingual

Workplace preferences

Remote only

Primarily works remotely

Groupama SA
Senior AI Engineer (Python | FastAPI | LLM | RAG)
BANKING AND INSURANCE
August 2025 - Today (10 months)
Paris, France
Development of a backend platform for intelligent simulation and calculation for car insurance, designed to be extensible to generative and advanced analytical AI uses.
Design and development of FastAPI APIs for heavy processing and complex business calculations.
Management of massive data volumes via Pydantic and asynchronous processing with Celery + Redis.
Deployment of services using LLMs for business analysis and automatic generation of summaries (discrepancies, anomalies, impacts).
Integration of LangChain: prompt chaining, business context management, JSON structuring.
Implementation of RAG pipelines: document indexing, vectorization, semantic search, enrichment of LLM responses.
Performance optimization: profiling, Pandas vectorization, I/O reduction.
Key results:
Efficient processing of +3.5M contracts across 12 entities.
Stable, scalable platform ready for generative AI in production.
FastAPI Pandas LLM Python PostgreSQL
Veolia
Lead Dev Python | FastAPI | Angular | GCP
ENERGY AND UTILITIES
March 2023 - June 2025 (2 years and 3 months)
Paris, France
Development of a multi-source supervision platform on GCP for large-scale data ingestion, processing, and exposure.
Data modeling & migration: design of business models with SQLAlchemy/Alembic (AlloyDB), migration pipelines between BigQuery, MongoDB, and AlloyDB, robust synchronization services (partitioning, batching, rollback).
Orchestration & automation: business workflows with Airflow (Composer) and Cloud Workflows, intelligent triggering via Cloud Scheduler, nightly pipelines for DMA and sector leak detection.
APIs & business services: development of secure REST APIs with FastAPI, automatic calculation of daily KPIs (consumption, pressure, reservoir level), Firestore emulator for local tests.
Architecture & best practices: standalone libraries for connectors (BigQuery, AlloyDB, MongoDB, InfluxDB), Dockerized deployment and IaC with Terraform, structured logging and multi-environment supervision.

Key results:
Efficient synchronization of 22 million lines in AlloyDB.
Significant optimization of calculation performance.
FastAPI Terraform Docker Google Cloud Platform (GCP) MongoDB
Cour des comptes
TechLead Python | Django | Angular
PUBLIC SECTOR
September 2019 - January 2023 (3 years and 4 months)
Paris, France
Development of control tools for financial jurisdictions in a critical and regulated environment.
Development of key functionalities: authorizations, dynamic forms, control lifecycle.
APIs & Backend: REST for repository and user account management, generation of advanced Excel exports (openpyxl).
Orchestration & automation: Airflow jobs, OpenID Connect authentication.
Architecture & refactoring: implementation of Clean Architecture, Python/SQL/NgRx audit and refactoring.
Research POC: Elasticsearch vs. Postgresql Vector.
Standalone Angular Libraries
Angular design system (Storybook, Material, Ng-Zorro).
Intelligent LDAP/Active Directory auto-completion.
Notifications/Pop-ups with NgRx, Jest.
DevOps & Industrialization
Jenkins → GitLab CI migration, complete CI/CD pipeline (Docker, SonarQube, Nexus, Ansible).
Load testing with Locust, backend test coverage 0 → 100%, frontend 80%.
Leadership & Methodology
Technical lead, mentoring, pair programming, code review.
Agile/Scrum team of 12 people.
Key results
Stabilization of a critical application in a regulated environment.
Performance optimization: removal of GraphQL, optimized REST APIs, Redis cache, reduced response times.
Adoption of Clean Architecture, isolation of use cases, introduction of ADRs and design patterns (Facade, Singleton, Repository…).
Industrialization of best practices: CI/CD, pre-commit, SonarQube, code conventions, security.
Implementation of technical documentation, changelog, and structured release process.
Proactive technical lead: risk alerts, functional challenges, backlog arbitration, and technical debt.
Python/Django Angular/Ngrx PostgreSQL Docker Jenkins