Dr. Gerrit T.

I build production-ready GeoAI and Deep Learning solutions for infrastructure, climate risk, and remote sensing applications — from research through to cloud deployment.

With a PhD in geomorphology (University of Göttingen) and 10+ years of applied data science experience, I combine rigorous domain expertise with end-to-end ML engineering. A recent reference project: an 18-month engagement with GEO DATA GmbH (May 2024 – November 2025), building a PyTorch-based CNN pipeline for automated infrastructure segmentation, deployed on Azure Machine Learning and Azure Data Factory under EU AI Act compliance, with MLOps practices and auxiliary ML tooling in LightGBM.

Core services:

- GeoAI and Computer Vision: deep-learning segmentation and object detection from aerial and satellite imagery

- Climate and flood risk modelling: hydrological modelling with geomorphological classification for site-level risk assessment

- Production ML engineering: end-to-end pipelines in PyTorch, Python, Azure ML and Azure Data Factory, with MLOps and EU AI Act compliance

- Geospatial analytics: PostGIS, rasterio, geopandas, QGIS/ArcGIS integration

- AI multi-agent systems for automated analysis workflows (PydanticAI)

Typical clients: organisations that work with geospatial data, remote sensing, or AI systems — across energy, infrastructure, insurance, real estate, environmental, and public sectors.

Available for consulting engagements, feasibility studies, and full implementation projects. A detailed project portfolio can be shared on request.