Freelancer profile translated to English.

Description

Looking for an expert to design and deploy an end-to-end RAG architecture in an enterprise setting? Build robust LLM pipelines for your GenAI projects? Integrate generative AI into your business workflows?

With nearly 10 years of experience in demanding environments (Sanofi, BNP Paribas CIB, Société Générale), I specialize in your Generative AI and Data Engineering projects, from scoping to production.

🚀 What I bring concretely:

Currently at Sanofi: Design and development of the complete RAG architecture for a pharmaceutical GenAI platform (parsing, chunking, embedding, vector store, retrieval, LLM generation). Azure OpenAI, AWS Bedrock, Pinecone, Weave/W&B observability.

Generative AI & RAG: Deployment of a documentary assistance RAG platform at BNP Paribas, reducing search time by 90%. LangChain, LangGraph, Vector DB, FastAPI.

Data Engineering: End-to-end AML & Fraud Detection pipelines processing 50M+ transactions/day. Spark, Kafka, Kubernetes, AWS.

Technical Leadership: Building and managing a team of 7+ people (Data Engineers, DevOps, BA). Code reviews, architectural decisions, mentoring.

ML in production: Fraud detection model (GBM/H2O) deployed in production at Société Générale.

📦 Technical Stack: Python, LangChain, LangGraph, Azure OpenAI, AWS Bedrock, Pinecone, FAISS, S3 Vectors, Scala, Spark, Kafka, FastAPI, Kubernetes, Docker, Terraform, Databricks, Weave/W&B, Snowflake. Forward Deployed Engineer, GenAI, FDE, API Integration, Python, Backend Development, LLM Orchestration

🎯 Sectors: Pharma, Banking, Finance, Insurance | Available | Île-de-France & Remote

Industry field of expertise

Languages

English
Native or bilingual
French
Native or bilingual

Workplace preferences

Can work on-site

Paris (up to 50km), Lyon (up to 50km), Lille (up to 50km), Nanterre (up to 50km), Bordeaux (up to 50km)

Sanofi Accelerator
Sanofi - Data & AI Engineer
PHARMACEUTICALS INDUSTRY
April 2026 - Today (4 months)
Paris, France
Context:
GenAI platform for automated generation of regulatory documents (Clinical Trial Documents) in the pharmaceutical industry. Critical production environment with strict traceability, security, and compliance requirements.
Achievements:

Design and development of the end-to-end RAG architecture: document parsing, chunking, embedding, vector store (Pinecone, S3 Vectors), retrieval, and LLM generation
Integration of LLM models in production: Azure OpenAI (GPT-4o), AWS Bedrock (Claude)
Observability architecture for LLM pipelines with Weave/W&B: step-by-step tracing for Data Science teams
Performance optimization: replaced FAISS with pre-calculated S3 Vectors, reducing costs by ~70%
Refactoring of the backend architecture towards DDD-light: resolved 12 audit findings
Writing technical specifications (16-section design doc) aligning Data Science, Data Engineering, and Backend
Multi-environment configuration (dev/test/prod) with Pinecone and EventBridge

Technical Stack:
Python 3.12 · FastAPI · AWS (Lambda, Step Functions, ECS, S3, Bedrock) · Azure OpenAI · LangChain · Pinecone · Weave/W&B · Terraform · Docker · GitHub Actions · Snowflake · Forward Deployed Engineer, GenAI, FDE, API Integration, Python, Backend Development, LLM Orchestration, NestJS · React · TypeScript
Langchain Retrieval-Augmented Generation (RAG) Python AWS artificial intelligence
BNP Paribas CIB
Senior Data & AI Engineer
BANKING AND INSURANCE
May 2022 - February 2026 (3 years and 9 months)
Pantin, France
Involvement in the Data Engineering and Generative AI projects of the IT Trade Finance team, focusing on AML (Anti-Money Laundering) and Fraud Detection.

📊 Data Project — AML & Fraud Detection Pipelines
Development of end-to-end pipelines processing millions of transactions: ETL, transformation, scoring, and alert generation.
→ Spark optimization (advanced tuning, data skew management)
→ Quantexa integration for relational graphs and contextual alert enrichment
→ Private cloud deployment with Kubernetes, Skaffold, Kustomize

👥 Setup and Structuring of a New Data Engineering Team
Leading the creation of a data team from scratch with 7+ members: defining needs, recruitment, onboarding, and skill development.
→ Creation and scaling of an offshore team in India (4 Data Engineers, 1 DevOps, 1 BA, 1 PO)
→ Implementation of development standards, architectural patterns, and best practices
→ Daily technical supervision: code reviews, architectural decisions, mentoring

🤖 Generative AI Project — Documentary Assistance RAG Platform
Design and deployment of a conversational platform to query all project documentation (Confluence, Jira, Elasticsearch, emails) in natural language.
→ 90% reduction in information search time for teams
→ Multi-source vectorization pipeline, vector database, LLM orchestration via LangChain with prompt engineering and optimized retrieval strategies
→ Python/FastAPI backend API, Kubernetes deployment

Stack: Python, LangChain, LangGraph, FastAPI, Elasticsearch, Vector DB, Scala, Spark, Kafka, Kubernetes, AWS, S3, Quantexa, ELK, RAG
Spark RAG Langchain Scala Python
Bedrock streaming
Senior Data engineer
PRESS AND MEDIA
January 2022 - May 2022 (4 months)
Lyon, France
Freelance mission within the A/B Testing team, on the M6+, RTL+ Hungary, and Videoland streaming platforms.

📊 Multi-platform Data Pipelines
Design and development of real-time and batch pipelines for experimentation and analytics across multiple international streaming platforms.
→ Ingestion of high volumes of user events via AWS Glue, EMR, and Athena
→ Scalable workflows with Spark and Databricks to ensure reliability of experimentation metrics
→ Infrastructure automation via Terraform and CI/CD pipelines (Jenkins, GitHub Actions)

Stack: AWS (Glue, EMR, Athena), Terraform, Python, Scala, Spark, Databricks, Airflow, Docker, Jenkins, GitHub Actions, Iceberg, dbt
Terraform Spark Python AWS Databricks