Hi, I'm
MSc in Physics and senior developer, eleven years into building production systems. I write the numerical and machine learning code in C# — most of it open source in CSNumerics, past 6,000 downloads on NuGet — and the applications that put it in front of people: a 3D exoplanet explorer with over 10,000 installs, and rain and flow analysis for Swedish wastewater networks. Earlier, AI and computer vision for industrial glass production.
Explore my work →I'm a physicist and developer based in Sweden, with an MSc in Physics from Linnaeus University and a decade of building production software. Since 2015 I've been part of Zenta in Växjö, building cloud-based business systems, integration platforms and digital tools for customers across many industries. My work bridges the gap between rigorous mathematical foundations and practical software engineering, creating tools that are both powerful and intuitive.
Beyond coding, I'm project coordinator for Mattecentrum in Växjö and chair of the local association, running the free drop-in maths help that gives students homework support at no cost. I'm fascinated by how we can make complex concepts accessible through technology — and by putting people in a room with someone who has time to explain.
A numerical library for C#, written from scratch in pure .NET: linear algebra, transforms, ODEs and integration; statistics, stochastic processes and Monte Carlo; supervised, unsupervised and sequence models (LSTM, TCN, 1D CNN) with grid search and cross-validation; and physics from classical mechanics to general relativity and quantum computing. It powers Exoplanet Hunter, a risk analysis platform and a quantum simulator.
// Detect exoplanet transits in a light curve var (X, y) = SequenceDataHelper .CreateWindows(lightCurve, 20, 1, 5); var result = SupervisedExperiment .For(X, y) .WithGrid(new PipelineGrid() .AddModel<CNN1DClassifier>(g => g .Add("Filters", 8) .Add("KernelSize", 5) .Add("Epochs", 150))) .WithCrossValidator( CrossValidatorConfig.KFold(folds: 3)) .Run(); // result.BestScore → 0.80+, cross-validated
Information pages and interactive 3D simulations for every known exoplanet and its host star, with live data from the Planetary Habitability Laboratory's exoplanet catalogue. Built as a React and TypeScript app with react-three-fiber for the orbital rendering, and shipped to Android through Capacitor from the same codebase.
Building and publishing my own scientific computing products: the CSNumerics library on NuGet, Unity 3D visualisation tools built on top of it, and applications on Google Play. The common thread is letting developers and researchers simulate, explore and visualise complex physics without building the numerics themselves.
Developing and maintaining a rain analysis and statistics library for processing measurement data from sewer networks. It identifies rain events to both Swedish and Danish standards, derives return periods, concentration time and runoff, calculates contributing area from three-day rain events, and produces KPI values from dilution ratio duration curves. It also turns raw sensor signals into flow — from pump cycles, and from water level via Manning's formula.
Led research and development on process optimization using artificial intelligence and machine learning. Project involved data collection systems, computer vision for bubble detection, and predictive modeling for glass quality improvement at industrial glassworks. Collaborated with glass manufacturers on optimization strategies.
Building and maintaining a comprehensive numerical library for C#, spanning numerical analysis, statistics, machine learning and physics. Published as NuGet packages with open source on GitHub, documentation at csnumerics.com and video tutorials walking developers and researchers through the theory behind the code.
Expert consultant in the research project behind Future City Flow, a collaboration between DHI, research institutes and Swedish municipalities. The platforms analyse rainfall and its effects on sewer systems, integrating GIS services to visualise hydrological data in map applications. Responsible for developing a data analysis tool that performs calculations on time series data to produce KPI values for measures on the pipe networks.
Developing digital products and services for customers across many industries — cloud-based business systems, integration platforms and web applications built with C# and .NET. Work spans agile product development, system integrations and UX-focused tooling.
Volunteer tutor at the free drop-in maths help run by Mattecentrum, the Swedish non-profit providing homework support to students nationwide, since 2013. Project coordinator and chair of the local association since 2021 — recruiting and scheduling volunteer tutors, and keeping the sessions running week after week. Featured in local press for contributions to accessible mathematics education — read more.
Built and still maintain an application for exploring astronomical data, now past 10,000 installs on Google Play alongside the web version. Rebuilt over the years from its original stack to React and react-three-fiber, with Capacitor packaging the same codebase for Android.
Five years of studies in the physical and mathematical sciences, with a strong computational emphasis: Physics 135 credits, Mathematics 82.5 credits, Computer Science 30 credits and Computer-aided design 7.5 credits.
Multi-year research project developing AI-driven systems for industrial glass production. Implemented data collection infrastructure, machine learning models for quality prediction, and computer vision for bubble detection. Results published through RISE and Vinnova — read the final report (PDF).
Vinnova-funded research project on urban water systems and smart city solutions, run by DHI together with research institutes and Swedish municipalities. Expert consultant on digital tools for water resource management and environmental monitoring, including GIS visualisation of hydrological data and KPI analysis of time series from the pipe networks.
CFD study of inlet box geometry for high pressure axial flow fans, using RANS turbulence modelling in COMSOL to reduce pressure losses and even out the velocity profile upstream of the impellers. Published through Linnaeus University — read the abridged thesis (PDF) or the full record on DiVA.
05. What's next?
I'm open to collaboration, interesting projects, and conversation. Whether it's about mathematics, numerical computing, or web development — feel free to reach out.
Send me an email