Data & analytics engineer · Energy, geospatial BI and forecasting
Portfolio · Blog · CV · LinkedIn
I build data pipelines and analytical applications for energy and infrastructure decisions. I'm interested in the whole path from a business question to a working tool: choosing the data, making assumptions explicit, and giving people a way to investigate the result.
My current work brings together geospatial BI, energy and renewables analytics, and reproducible Python and SQL workflows. I'm also developing my forecasting and data-platform skills.
Where should a charging operator expand next? I built NRW Charging Atlas to support early siting decisions for operators planning charging infrastructure for electric vehicles (EV) across all 53 districts in North Rhine-Westphalia (NRW), Germany.
You can screen districts for a rollout, inspect what drives their priority rankings and test proposed stations against current coverage. I pushed spatial joins and scoring into PostGIS, with Python for data validation and loading, GeoServer for map and feature services, and GeoNode for dataset discovery and metadata.
Explore the code · Read the project story · Try the visual walkthrough · Open the slides
Look inside the GeoNode dataset catalogue
The catalogue keeps published layers and their source information together. Proposed stations start empty; district layers share boundaries, so their previews can look alike even though their values differ.
Map previews: © EuroGeographics and © OpenStreetMap contributors.
agri-weather-yield-drivers combines weather, soil and yield data into explainable risk signals. The work includes checks on spatial joins, coverage and baseline choices, with a reporting mart and risk-zone GeoJSON as outputs.
bi-python-uv-project-scaffolder creates a starting structure for BI projects using Python and uv. It supports the practical setup work behind repeatable analytical pipelines.
- NRW Charging Atlas: comparing charging coverage with GeoNode and PostGIS: the siting question, scoring model and data pipeline.
- A quick tour of NRW Charging Atlas: start the application, try a proposed station and explore the data services.
- Crude benchmarks: API, sulfur and interactive charts: explore crude quality through interactive charts.
Python · SQL · PostgreSQL / PostGIS · DuckDB · GeoNode · GeoServer · pandas · NumPy · scikit-learn · PyTorch · Power BI · Docker · AWS · TypeScript · React
Languages across my public repositories
I'm open to conversations and collaboration around energy analytics, geospatial BI and forecasting. Find me on LinkedIn or explore my portfolio.