About Us
From research to production AI.
GeoLambda GmbH is a research, development and consultancy company at the intersection of geoinformatics and applied AI. We develop methods and tools that go beyond conventional GIS workflows and off-the-shelf models. Founded in 2025 by Dr. Gerrit Tombrink (Dr. rer. nat. in Geomorphology, University of Göttingen).
What makes us different
GeoAI at the core – We combine state-of-the-art computer vision, large language models and purpose-built AI agents to automate image classification, object detection and geospatial ETL. Our satellite segmentation system classifies alpine land surfaces with over 95% accuracy, and a recent client project — a deep learning tool and pipeline for geodata segmentation — was delivered end-to-end on Azure with GPU-scaled inference.
Built-in AI safety & EU AI Act readiness – Our internal Secure-AI checklist (bias audit, explainability, adversarial tests) is informed by the Commission's JRC guidance on harmonised AI Act standards [JRC 139430], the security specifications of ETSI TC-SAI and the independent safety recommendations of the Future of Life Institute. We map every system to the EU AI Act risk classes.
Grounded in field research – Our work builds on five research expeditions across the Himalayas, the Karakoram and the Andes, and has been published in peer-reviewed journals (Springer, Copernicus). This earth science foundation shapes how we approach climate risk, natural hazard analysis and environmental monitoring.
GeoLambda pairs GeoAI research, field-tested earth science and EU AI Act-aligned engineering to turn your location data into actionable insight — responsibly and fast.
Founder & Expertise
Meet the founder behind GeoLambda.
Dr. Gerrit Tombrink
Founder, CEO & Data ScientistGerrit Tombrink holds a doctorate (Dr. rer. nat.) in high mountain geomorphology from the University of Göttingen, where he researched flood dynamics and proglacial stream systems in the Himalayas. His doctoral work under Prof. Dr. Matthias Kuhle included five research expeditions across the Himalayas, the Karakoram and the Andes, funded through competitively awarded research grants. He also holds a Postgraduate Certificate in Geoinformatics (UNIGIS) from the University of Salzburg. Before founding GeoLambda GmbH, he spent three years as a Geospatial Data Scientist. In this role, he developed a deep learning tool (PyTorch/fastai, ResNet50/U-Net) for satellite image segmentation that classifies alpine land surfaces with over 95% accuracy. Afterwards, he worked as a freelance data science consultant for four years, delivering projects for clients including GEO DATA GmbH, the Max Planck Institute, geoinformation software manufacturers and energy distribution network operators. His research has been published in, among others, the Journal of Mountain Science (Springer), the E&G Quaternary Science Journal, and as a Springer conference chapter (UIS 2024).
Selected Publications
A. Abecker, M. Budde, F. Fuchs-Kittowski, J. Großmann, W. Koch, J. Lachowitzer, E. Rodner, H. Rudolf, P. Schulze, G. Tombrink, M. Zemann (2025). Herausforderungen und Ansätze zu einer Infrastruktur für die breite Nutzung von Machine-Learning-Verfahren in der Umweltverwaltung. In: Umweltinformationssysteme (UIS 2024), pp. 113–133. Springer. DOI: 10.1007/978-3-658-46394-6_8
G. Tombrink (2018). Proglacial streams and their chronology in the glacier forefields of the Himalayas. E&G Quaternary Science Journal, 67, 33–36. Copernicus Publications. DOI: 10.5194/egqsj-67-33-2018
G. Tombrink (2018). Der glazifluviale Formenschatz im Gletschervorfeld des Himalaya und der Versuch einer relativ-zeitlichen Einordnung. Dissertation, Georg-August-Universität Göttingen. DOI: 10.53846/goediss-6660
G. Tombrink (2017). Flood events and their effects in a Himalayan mountain river: Geomorphological examples from the Buri Gandaki Valley, Nepal. Journal of Mountain Science, 14(7), 1303–1316. Springer. DOI: 10.1007/s11629-016-4154-5
Project Highlights
Selected projects delivered by our founder — now continued through GeoLambda GmbH.
Deep learning methods (AI) for geodata analysis
05/2024 - 11/2025AI consulting, development, and training of neural networks for segmenting construction trenches and trench centre lines using Python, Azure Machine Learning, Azure Data Factory, and Azure DevOps.
MLOps, AutoML and web development
07/2023 - 01/2024Evaluation and testing of AutoML and MLOps methods with Python, development of a web survey with the Python library Flask, including database transfer.
Data Wrangling and ML App Development
10/2022 - 12/2022Web Scraping and Data Wrangling of spatial datasets with Python, development of an (end-to-end) ML app and dashboards for training/validation of datasets with Python.
Database analytics, mapping and visualisation
07/2021 - 06/2022Implementation of ETL processes to transform PostgreSQL/PostGIS data into an ArangoDB, creation of tools (e.g. visualisation, validation and anonymisation) and dashboards with Python.
Flood events and their effects in a Himalayan mountain river
2015 - 2017Investigations of the Buri Gandaki river (Nepal) system. Research examines flood events and related human interactions in the northwestern Himalayan Buri Gandaki Valley.
Multi-Agent Research Automation
01/2026An open-source multi-agent system that autonomously researches and identifies domain experts for conferences and panels. Agents collaboratively query the web, extract structured data and rank candidates — demonstrating agentic AI architecture with real-world applicability to research automation and lead intelligence.
Ready to transform your geospatial data?
We reply within one business day.