matrixise - Overview

Hi there 👋 I'm Stéphane Wirtel

Senior Python Expert | CPython Core Developer

🐍 CPython Core Developer since 2019
🎤 International conference speaker (PyCon, EuroPython, FOSDEM)
📝 Technical writer & blogger at wirtel.be
🥋 2nd DAN Karate JKA
🏢 Founder of MGX.IO - Python consulting

📚 Latest Blog Posts

➡️ Read more on my blog

🚀 Featured Projects

CPython Core Developer

Contributing to the Python programming language - bug triage, documentation improvements since 2019.

PythonFOSDEM

Co-Founder and co-organizer of PythonFOSDEM, a community event that has grown from 80 to 800+ attendees, bringing Python enthusiasts together at FOSDEM.

Python Ireland

PyCon Ireland co-organizer (2020-2025) & python.ie maintainer — 11 years serving Ireland's Python community through code, talks, and event organization.

dsmtpd

SMTP debugging tool for Python developers - a project that evolved from a simple script to version 1.0.0 over 12 years.

🎤 Conference Talks

I've delivered 9 major talks across Europe and North America at conferences including:

PyCon Ireland - Python and open source development
PyCon Canada - CPython internals and contribution workflows
EuroPython - Advanced Python topics
FOSDEM - Community building and Python ecosystem

🛠️ Tech Stack & Skills

Areas of Expertise:

CPython internals and core development
Machine Learning & Data Science
Educational data mining and predictive analytics
Web frameworks (Django, FastAPI, Wagtail)
Database design and PostgreSQL optimization
Web3 and blockchain development (Solidity, smart contracts)
Financial systems and risk modeling
Open source community leadership
DevOps & Infrastructure automation (Ansible, Docker, Traefik)
System administration and large-scale Debian migrations

🤖 Machine Learning & AI

Classical ML & Data Science

Recent Projects:

📊 Educational Predictive Analytics - ML model to identify and support students at risk of dropout
🔄 End-to-End ML Pipelines - From data exploration to production deployment with testing and CI/CD
📈 Feature Engineering - Creating business-driven features (success rates, performance trends, segmentation)

ML/Data Science Stack:

Core ML: scikit-learn (pipelines, preprocessing), pandas, numpy
Visualization: matplotlib, seaborn, plotly
Tools: Jupyter notebooks
MLOps: Reproducible builds, model versioning with joblib, CI/CD for ML

Deep Learning & NLP

DiagnosIA - Medical Diagnosis Classification System

Built a production-ready medical diagnosis system using Bio_ClinicalBERT for symptom-to-disease classification with 6 disease classes.

Key Achievements:

Implemented transfer learning with Bio_ClinicalBERT fine-tuning
Designed intelligent medical data augmentation (shuffle, drop, templates) preserving clinical meaning
Achieved production deployment with 4 user interfaces (CLI, REST API, Streamlit, Chainlit)
Architected BERT + LLM hybrid system: BERT for prediction, LLM for clinical summaries
Multi-backend LLM support (Ollama local, Lightning.AI cloud)

Deep Learning Stack:

Core Skills:

Transfer learning & model fine-tuning
Text classification with transformers (BERT family)
Production ML pipelines (training, evaluation, deployment)
Multi-device optimization (CUDA, MPS, CPU)
LLM integration and prompt engineering
ML application development (API, Web UI, Chat)

📊 GitHub Stats

📫 Connect With Me

🎯 Currently

📖 Writing and relaunching my Python 3.14 book
🐍 Contributing to CPython and maintaining the Developer Guide
🎪 Organizing PythonFOSDEM community events
💼 Available for Python consulting and speaking opportunities
🏗️ Building infrastructure automation with Ansible and Docker
🌱 Exploring the intersection of Python and blockchain technologies
🤖 Building Deep Learning applications with transformers and LLMs (see DiagnosIA project)

Last updated: April 2026