Pierre Moreau

PhD in Machine Learning · AI Healthcare Engineer

AI researcher specialized in healthcare applications — multimodal learning, clinical risk assessment, generative AI. Founder of Noovate-lab, an AI-first studio. Available for AI consulting in the healthcare industry.

Get in Touch View Projects CV (PDF)
0
peer-reviewed publications
0
years of software development
0
PhD defended (computer science)

About Me

I am an AI & Machine Learning researcher with a PhD in computer science (Nov 2023), specialized in healthcare applications. My doctoral research, conducted in partnership with Amiens University Hospital, focused on motion analysis using artificial intelligence — particularly applied to Parkinson’s disease.

I am currently a post-doctoral researcher at the University of Picardie Jules Verne, working on the RAUC project: an optimized care pathway for digestive surgical emergencies, using AI to predict post-operative complications. Previously, I led the AI & Data Science workstream at Capgemini Engineering on a generative AI medical assistant demonstrator, coordinating a team of 15–20 people.

In parallel, I’m founding Noovate-lab, an AI-first studio building practical products for healthcare and SMBs. I bring deep expertise in LLM-based systems, multimodal RAG, computer vision (medical imaging), explainable AI, and production deployment of ML pipelines.

Amiens, France
moreau.p.02@gmail.com
+33 6 11 80 58 32
LinkedIn Profile
GitHub Profile

Currently building

last updated · may 2026

research

Post-doc RAUC at UPJV

Predicting post-operative complications for digestive surgical emergencies, in partnership with Amiens University Hospital.

product

Solis V1.0 launch

Wrapping the AI ERP for SMBs (devis, factures, dictée vocale, OCR) and recruiting 10 beta testers among Amiens-area artisans.

product

Boussole MVP — Q4 2026

AI assistant for French citizen benefits (60+ via OpenFisca) and legal recourses. Three-pillar architecture, MVP shipping in November.

consulting

Available from Q3 2026

Open to AI consulting missions in healthcare (audits, MVPs, LLM integrations in clinical workflows). 5–15 day formats.

Get in touch →

Professional Experience

09/2025 — present

Post-Doctoral Researcher in AI

University of Picardie Jules Verne — Amiens, France

  • Project RAUC: optimized care pathway for digestive surgical emergencies, integrating AI and e-health
  • Predicting post-operative complications from clinical data
  • Partnership with Amiens University Hospital (CHU)
03/2024 — 05/2025

Data & AI Engineer — AI & Data Science Lead, Healthcare

Capgemini Engineering — Paris, France

  • Led the design and development of a GenAI-powered medical assistant demonstrator with multimodal workflows
  • Designed NLP pipelines: OCR for blood test analysis, Whisper speech-to-text, automated keyword extraction
  • Fine-tuned LLMs (OpenAI, Mistral) with PEFT / LoRA for clinical summarization and audio transcription
  • Built a multimodal RAG pipeline with LangChain to extract and synthesize information from structured and unstructured medical data
  • Applied research on placenta segmentation with Transformer architectures (UNetR, SwinUNet) for prenatal imaging
  • Coordinated AI researchers, data scientists and clinicians (team of 15–20 people)
2022 — 2023

Teaching & Research Associate (ATER)

University of Picardie Jules Verne — Amiens, France

  • Lectures and labs for 1st, 2nd and 3rd-year undergraduates (PHP, Bash, C, advanced programming)
  • Course design on machine learning fundamentals
2019 — 2023

PhD Research — Human Movement Modeling & Analysis

University of Picardie Jules Verne & Amiens University Hospital

  • Built a comprehensive human-motion database for quantitative analysis and model training
  • Data analysis with PCA, Dynamic Time Warping (DTW) and correlation
  • Genetic algorithms for parameter tuning and feature selection
  • Deep-learning models (RNN, LSTM, CNN, residual networks) and SVM for classification
  • Interpretable biomechanical representations via matrix-based dynamic models
  • PhD defended November 2023
2017 — 2019

Software Engineer (apprenticeship then full-time)

LV Informatique — Amiens, France

  • Developed and maintained the Lynéade / Lynéa2 ERP (WinDev): accounting, purchasing, sales, inventory, after-sales
  • SQL database administration and performance tuning
  • Customer-driven feature delivery — direct foundation for my current ERP product (Solis)

Featured Projects

A mix of Noovate-lab studio products (in production / in development) and research & professional projects.

Noovate-lab studio

Solis — AI-powered ERP for SMBs

Full-stack ERP for French artisans / TPE, in late development: quotes, invoices, credit notes, deposits, e-invoicing (Factur-X), Stripe billing, OCR scan, voice dictation (Whisper), conversational AI assistant. RGPD audit, design system « Atelier ».

Django Next.js 15 Claude API Whisper Stripe Factur-X
View Live Demo

Boussole — AI assistant for benefits & legal recourse

An AI assistant that aggregates all French benefits (60+ via OpenFisca, ~1500 local aids) and legal recourses in one place. Three pillars: citizen benefits, sector-specific aids (doctors, teachers, civil servants), and legal litigation (caution disputes, etc.). MVP in development.

Next.js 15 Supabase Claude API RAG OpenFisca

Vesta — Radiology planning

Web app generating monthly schedules for the Radiology Department at Amiens University Hospital — 5 specialties, 6-week rotation, manual locks with auto-regeneration. Client-side only (React + Vite), persisted in localStorage.

React Vite TypeScript Hospital deployment

Research & professional projects

Medical AI Assistant (Capgemini)

GenAI-powered medical assistant demonstrator with multimodal processing: OCR for blood test analysis, Whisper speech-to-text, fine-tuned LLMs for clinical summarization, multimodal RAG pipeline.

Python LangChain PEFT / LoRA RAG

Placenta Segmentation (3D Doppler)

Transformer-based segmentation (UNetR, SwinUNet) for earlier preeclampsia detection on 3D Doppler images. Submitted to CMIG 2025 (co-author).

PyTorch UNetR SwinUNet Medical Imaging

Human Movement Analysis (PhD)

Comprehensive motion modeling for Parkinson’s disease and human activity recognition. Deep-learning architectures (RNN, LSTM, CNN), genetic algorithms, interpretable matrix-based representations. 3 publications.

RNN / LSTM PCA / DTW Genetic Algorithms SVM

Studio Noovate-lab

Noovate-lab is an AI-first studio I’m founding in 2026, building practical products and offering consulting at the intersection of healthcare, AI, and small-business software.

Products

SaaS we ship and own. Solis (AI ERP for SMBs, late development), Boussole (benefits & recourse assistant, MVP Q4 2026), Vesta (hospital planning, in production at CHU Amiens).

AI Consulting — Healthcare focus

Audits, MVPs, integrations of LLMs into clinical workflows. Drawing on my PhD + post-doc + Capgemini Healthcare lead experience. Typical missions: 5–15 days, scoping and prototyping.

Applied research

Bridging academic ML (publications, post-doc RAUC) with production-grade engineering. Computer vision on medical imaging, RAG over clinical corpora, explainable AI.

Available for AI consulting missions starting Q3 2026.

Discuss a project

Publications

2025 Submitted

Towards Better Placenta Segmentation for Earlier Preeclampsia Detection on 3D Doppler Images

Computerized Medical Imaging and Graphics (CMIG)

Charlotte Alliod, Emilien Micard, Pierre Moreau, Léo Challier, Georgia Martensen, Hocine Kadi, Solène Le Bars, Olivier Morel, Marc Blanchon, Marine Beaumont

2023 Published

Two-branch neural network using two data types for human activity recognition

IEEE Sensors Journal

Pierre Moreau, David Durand, Jérôme Bosche, Michel Lefranc

2021 Published

A motion recognition technique based on linear matrix representation to improve Parkinson’s disease treatments

9th International Conference on Systems and Control (ICSC)

Pierre Moreau, David Durand, Jérôme Bosche, Michel Lefranc

2020 Published

A motion recognition algorithm using polytopic modeling

7th International Conference on Control, Decision and Information Technologies (CoDIT)

Pierre Moreau, David Durand, Jérôme Bosche, Michel Lefranc

Technical Skills

ML & Deep Learning

PyTorch TensorFlow Keras scikit-learn XGBoost Hugging Face CNN / RNN / LSTM Transformers Reinforcement Learning Genetic Algorithms

GenAI & LLMs

Claude API OpenAI LangChain Multimodal RAG PEFT / LoRA Fine-tuning Prompt Engineering Vector DB

Computer Vision

UNetR SwinUNet OpenCV PIL / Pillow Medical Imaging Segmentation

NLP & Speech

NLTK Whisper (STT) OCR Tokenization Embeddings

Data Science & XAI

NumPy Pandas SciPy Matplotlib Seaborn PCA DTW SHAP LIME ANOVA

Backend & MLOps

FastAPI Django Flask Next.js 15 Supabase PostgreSQL Docker Stripe

Languages & Tools

Python SQL JavaScript / TypeScript Bash HTML / CSS Git Claude Code / Cursor Jupyter

AI Ethics & Compliance

RGPD / GDPR EU AI Act Bias / Fairness Explainable AI Quantization / Pruning

Languages

French — Native English — Upper Intermediate (B2)

Get In Touch

Let's Connect

I'm always interested in discussing new opportunities in AI and machine learning, particularly in healthcare applications. Feel free to reach out!

moreau.p.02@gmail.com LinkedIn Profile +33 6 11 80 58 32 GitHub Profile