Applied AI / ML Engineer · Toulouse

Julien Rabault

I build RAG over 100k documents, LLM agents, and Deep Learning trained on supercomputers.

Resume PDF Get in touch →
About

4 years at CNRS training models for researchers in weather, astrophysics, ecology and nanomaterials. Multi-GPU on Jean Zay, libraries still running in prod at Météo France, 2 peer-reviewed publications.

Now at Berger-Levrault, I build Athena: business users ask questions about ~100k technical documents, RAG and MCP agents find the answer and cite their sources.

3 packages on PyPI. When I'm not coding, I read sci-fi and hike the Pyrenees. Based in Toulouse, open to Paris / Montreal.

Experience

AI Engineer

Berger-Levrault

Jan. 2026 – present · Toulouse

AI R&D team · 12 people

Athena is Berger-Levrault's agentic platform: multi-agent, routing, connected to documents and business APIs for all clients (local government, industry, maintenance).
Deployment in progress, ~30 pilot users. Langfuse observability, client workshops, iterating on field feedback.

4 key contributions:

Athena · multi-agent architecture

Context: build the agentic foundation for all BL clients (local government, industry, maintenance).
Role: multi-agent architecture design (LangGraph), question-type routing, RAG + MCP agent orchestration, response sourcing integration.
Result: platform in production, source-grounded and actionable answers across multiple business domains.

RAGMCPLangGraphAgent Routing

Content-extractor · OCR/PDF/DOCX

Context: initial service was a basic image extractor and hard to extend.
Role: full OCR/images/PDF/DOCX rebuild, async batching (Celery + Mistral batch API), factory/registry architecture to add extractors/chunkers quickly.
Result: -50% extraction costs.

OCRCeleryMistral BatchFactory/RegistryPython

Airflow pipelines

Context: document ingestion per client account.
Role: took over existing setup and evolved it with the team for ingestion flows (PDFs, manuals, work orders, machine docs).
Result: 5 operational DAGs (1 per client).

AirflowDAGPDF/DOCXIngestionChunking

MCP Builder

Context: rapidly connect BL's internal APIs to agents.
Role: an LLM analyzes OpenAPI specs to pre-process endpoints (grouping, masking, description generation), then human-in-the-loop review to validate routes and business context.
Result: 120+ internal APIs mapped, with progressive runtime integration.

OpenAPIMCPHuman-in-the-loopAPI Integration

Machine Learning Engineer

CNRS · National AI Research Program (PNRIA)

Dec. 2021 – Jan. 2026 · Toulouse

Part of the PNRIA network, I worked with research teams across France on their AI problems.
Always 2 projects running in parallel (6-12 months each), delivered to partners like Météo France, CNES, CEA, INEE.
Training and fine-tuning on Jean Zay (multi-GPU DDP, up to 8 GPUs, Slurm). PyTorch profiling.

Selected proof points:

GENS / MetScore · Météo France

Context: evaluate weather models in production, multi-GPU optimization and fine-tuning of a diffusion model.
Role: designed MetScore (YAML configuration + Python library) and trained a diffusion model (DDPM) in PyTorch for weather-image generation on Jean Zay.
Result: library still in production + diffusion POC at -20% compute with equivalent quality.
Co-author on AMS 2025 paper

PyTorchDDPMHPCPython

DeepFaune · CNRS/INEE

Context: wildlife recognition on camera traps (1.5M images).
Role: fine-tuning YOLOv5 on custom dataset (1.5M images, 24 classes), multi-GPU training, class-imbalance handling, precision/speed optimization for CPU deployment.
Result: 93% on 24 species, 3× faster.
Peer-reviewed publication

Fine-tuningYOLOv5PyTorchMulti-GPUComputer Vision

Other contributions: AUTOFILL (CEA, PairVAE, MAE 0.98), BIGSF (CNES, tech lead and architecture refactor), Introduction to LLMs training (3h, ~25 CNRS PhD students/researchers).

Software Engineer

Agileo Automation (apprenticeship)

Aug. 2020 – Sept. 2021 · Montauban

Supervision and control framework for robotic machinery in semiconductor manufacturing. Object-oriented architecture, C#, HMI. Team of 5 engineers, Agile/Scrum.

Education

M.Sc. · Artificial Intelligence & Pattern Recognition (IARF)

Université Paul Sabatier Toulouse III · IRIT

Specialization in Deep Learning, Computer Vision, NLP. IRIT lab (Toulouse Computer Science Research Institute).

2019 – 2021

B.Sc. Computer Science

Université Paul Sabatier Toulouse III

2016 – 2019

Languages

French · native English · fluent (B2)
Open source projects 04 projects ·
LLMock PyPI

LLM mock server to test retries, fallbacks and rate limiting, zero tokens spent

DDPM-weather Open Source

Denoising diffusion probabilistic model for weather imagery · −20% resource usage

uniqpath PyPI

Python utility to generate unique file paths: incremental suffixes, UUIDs, timestamps

BananaML Production

End-to-end ML pipeline deployed on AWS · Computer Vision + REST API

Publication
Skills

Agentic AI & RAG

RAG / GraphRAG Multi-agent Systems MCP Protocol Prompt Engineering Structured Outputs Embeddings LLM Routing Fine-tuning

Deep Learning & Vision

PyTorch Transformers Computer Vision Diffusion Models NLP CNNs / U-Net / YOLO

MLOps & Infrastructure

Docker AWS CI/CD MLFlow Airflow Kubernetes Celery Langfuse HPC / Slurm

Engineering & Frameworks

Python FastAPI LangChain / LangGraph HuggingFace Qdrant / pgvector Mistral / OpenAI API Git SOLID / Architecture

Contact

Let's work together

Available for new opportunities. Toulouse, Paris or Montreal. Remote OK.

Domains: GenAI · RAG · Multi-agent Systems · MLOps · Computer Vision

Download my CV

Julien Rabault · Toulouse, France ·