LC-MS specialist with over two decades of hands-on mass spectrometry experience, and a self-taught programmer who builds tools to make lab data easier to trust and use.
This is where I keep personal projects, learning notes and experiments. Everything here uses public or synthetic data.
| Repository | What's inside |
|---|---|
| data-science | Python notebooks and scripts for data analysis: PCA from scratch, machine learning with scikit-learn, signal processing and public market data, plus a bilingual (French/English) learning-with-kids series in Python |
| lab-informatics | Readers, validators and writers for open laboratory data standards (JCAMP-DX and mzML), built on synthetic data, with unit tests running in CI on Python 3.9 to 3.15 |
| automation-workflow | Reproducible workflow patterns: Docker environments, scheduled containerised jobs on Windows, Docker cleanup notes, API clients, PDF automation and a Raspberry Pi setup runbook |
| arduino-showcase | Embedded prototyping: sensors, a multichannel voltage logger, intermediate-level concept units (state machines, analog input), and a bilingual (French/English) learning-with-kids series in Arduino |
| network-infrastructure | Practical network troubleshooting and infrastructure notes: a Windows diagnostics toolkit, CLI reference, networking fundamentals and a disk/storage guide |
| quality-engineering | Testing as a discipline: functional, negative, boundary and OpenAPI contract tests for a small local API service, with a test plan and CI |
- lab-informatics and quality-engineering: the code, tests and documentation were written by an AI assistant (Claude, through Claude Code) under my direction. I set the scope and requirements and publish everything by hand. These two repositories are also my learning material: open data standards and CI are new to me, and I am learning them by reading, running and extending this work.
- My earlier material in the other repositories (notebooks, scripts, prototypes, projects and course work that existed before this portfolio): the code is my own. I used the assistant to organise it, to help debug and optimise it, and to write documentation.
- New material built with the assistant since late September 2026 (for example the learning-with-kids lessons, generated from my own project code and course homework): written with the assistant from my own plans and material.
Working with AI, and checking it. I work in separate roles: one chat plans, debugs and writes instructions, the assistant executes, and the output comes back to me for review. What makes this trustworthy is the verification around it: automated tests, CI on several Python versions, a requirements-to-tests traceability matrix for the mzML tools, and test plans that list known gaps. I have read most of what is published and keep reading the rest; the checks exist for what reading alone cannot catch. A few commits carry the assistant's co-author trailer.
Last updated: 2026-11-10.
Python · Pandas · NumPy · SciPy · scikit-learn · TensorFlow · Jupyter · Docker · C/C++ · Arduino