Work with me
I’m a freelance Software & Data Engineer in Ghent, Belgium, open to new assignments.
- Availability: 3–4 days a week, hybrid (Ghent / Flanders).
- Languages: Dutch (native), French (fluent), English (fluent).
- Set-up: eenmanszaak, BTW-registered.
📄 CV (PDF) (web version) • ✉️ tomas.fiers@gmail.com
Software & data engineering
I’m a data scientist and full-stack developer who builds both the analysis and the product around it. Seven years of data science and scientific computing in research (machine learning, statistical modelling, signal processing in Python and Julia), plus production full-stack work: a work-order app for technicians with ERP integrations (TypeScript, React Native, NestJS, Kubernetes), and a usage-analytics system at DataCamp (Node.js, Angular, Redis, PostgreSQL). Burgerlijk ingenieur (signal processing) with a PhD. I care about correctness, clear communication, and good user interfaces, and like taking a problem from vague idea to a working, tested tool.
Skills
- Data science & machine learning — Python (NumPy, SciPy, pandas, scikit-learn, PyTorch/TensorFlow, matplotlib), lots of Jupyter, Julia, R, SQL, MATLAB, C. Bayesian statistics, neural networks and deep learning, data visualisation.
- Signal processing — time series analysis, filter design, Fourier/wavelet analysis, real-time detection, simulation.
- Software & web — TypeScript/JavaScript, React & React Native (Expo, EAS builds, OTA updates), Remix, Angular, Node.js/NestJS, Java, PHP, Kotlin (Android), HTML5/CSS. RESTful APIs, OAuth, webhooks, microservices.
- Data & infra — PostgreSQL/Prisma, Redis & Bull job queues, S3-compatible storage, Airflow, Docker, Kubernetes, GitHub Actions CI/CD, Sentry, HPC.
- Practice — Git, Linux, testing (pytest, Jest), functional analysis & specs, customer contact.
- Scientific communication & data visualization — clear, carefully designed figures and documents. (PhD thesis with margin notes, and figures linked to their source Jupyter notebooks for reproducibility).
- UX research & design — user interviews.
- AI-assisted development — spec-driven coding with agents (Claude Code, GitHub Copilot), and orchestrating multiple agents.
Recent work
Workshaker, for Softwareshaker (NL) — 2025 – 2026. Full-stack developer on a work-order app for technicians, used by 400+ companies, integrated with the Robaws ERP. Stack: React Native/Expo, Remix, NestJS, PostgreSQL/Prisma, Kubernetes.
- Features I built, from spec to production: a form builder with PDF export, ERP custom fields, stock movements, an hours overview, a scheduled-maintenance module, and time-off requests.
- Integration with the Robaws API (OAuth, webhooks, caching). Reverse-engineered undocumented API behaviour, and wrote feature requests to the vendor. Made the ERP data sync up to 4× faster with batching.
- A backward-compatible production data migration across all customer organisations.
- I did the deploys and monitoring, and investigated customer incidents (Sentry, kubectl, production queries).
WooCommerce → Robaws ERP integration — a WordPress/PHP plugin plus a Node service, with queue-based webhooks and auto-updates. Live at several customer webshops.
Mic Atlas (side project) — a microphone-comparison web app.
Before that: research
- Postdoc, neuroinformatics — Ghent University, 2024. Bayesian inverse-model estimation on large, heterogeneous brain-imaging datasets (many scan modalities). Python data pipelines and analyses, lots of notebooks. International collaboration in the European EBRAINS project.
- PhD, computational neuroscience — University of Nottingham, 2020–2024. Thesis: From Voltage to Wiring: Synaptic connectivity inference from neural voltage recordings (PDF), with ~100 research notebooks behind its figures. Statistical and machine-learning methods to infer neural connectivity from voltage-imaging data (Python, Julia). Built and performance-optimised a spiking neural network simulation library in Julia.
- COMOB — an open, massively collaborative research project led by Dan Goodman (Imperial College London): I set up its website and notebook infrastructure. Resulting paper: Spiking Neural Network Models of Interaural Time Difference Extraction via a Massively Collaborative Process (eNeuro, 2025).
- Research engineer, signal processing — Neuro-Electronics Research Flanders (imec, VIB), 2019. Real-time detection of hippocampal sharp-wave ripples with spatiotemporal filtering and recurrent neural networks.
- Software engineering internship — DataCamp, Leuven, 2016. A full-stack usage-analytics system for DataCamp’s live learning platform: it tracks how much time users spend in the app, collects the events in Redis, and aggregates them into PostgreSQL (Node.js, Angular, microservices). It was still in production at least until 2018, when DataCamp had 3 million users.
- Other internships — Byteflies (motion-sensor data for a medical wearable), Fluves (business plan for an offshore-wind joint venture).
Currently preparing for the ‘Microsoft Certified: Azure AI App & Agent Developer (AI-103)’ exam.