About Us

From research to production AI.

GeoLambda Lab

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 fuse state-of-the-art computer vision, large-language models and purpose-built AI agents to automate image classification, object detection and geospatial ETL — workflows that once needed whole analyst teams. Our satellite segmentation systems classify alpine land surfaces with over 95% accuracy.

Built-in AI safety & EU-AI-Act readiness – Our Secure-AI checklist (bias audit, explainability, adversarial tests) follows 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.

Lean, science-driven delivery – We run the build → measure → learn loop from the Lean-Startup playbook, shipping minimum-viable products in weeks, not quarters. The approach cuts waste and boosts product-market fit. Our most recent client project — a deep learning tool and pipeline for geodata segmentation — was delivered end-to-end on Azure with GPU-scaled inference.

Grounded in field research – Our work builds on five research expeditions across the Himalayas, the Karakorum 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, lean product thinking and secure AI to turn your location data into actionable insight — responsibly and fast.

Founder & Expertise

Meet the founder and visionary behind GeoLambda's AI-powered solutions.

Dr. Gerrit Tombrink

Dr. Gerrit Tombrink

Founder, CEO & Data Scientist

Gerrit 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 Karakorum and the Andes, funded through competitively awarded research grants. He also holds a degree in geoinformatics from Paris Lodron University Salzburg (UNIGIS). 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

Before founding GeoLambda GmbH, Dr. Tombrink established his expertise through key projects with leading institutions and enterprises:

Deep learning methods (AI) for geodata analysis

05/2024 - 11/2025
Client: GEO DATA GmbH

AI 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.

Python PyTorch Azure ML Azure Data Factory Docker MLOps
Read More

MLOps, AutoML and web development

07/2023 - 01/2024
Client: Manufacturer of geoinformation software

Evaluation and testing of AutoML and MLOps methods with Python, development of a web survey with the Python library Flask, including database transfer.

Python JavaScript HTML Pandas MLflow Flask
Read More

Data Wrangling and ML App Development

10/2022 - 12/2022
Client: Max-Planck-Institute and Helmholtz-Institute

Web 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.

Python Pandas NumPy Streamlit Plotly AWS
Read More

Database analytics, mapping and visualization

07/2021 - 06/2022
Client: Distribution system operator / Energy provider

Implementation of ETL processes to transform PostgreSQL/PostGIS data into an ArangoDB, creation of tools (e.g. visualisation, validation and anonymisation) and dashboards with Python.

Python PostgreSQL PostGIS ArangoDB NetworkX Bokeh
Read More

Flood events and their effects in a Himalayan mountain river

2015 - 2017
Type: Independent research project

Investigations of the Buri Gandaki river (Nepal) system. Research examines flood events and related human interactions in the northwestern Himalayan Buri Gandaki Valley.

Lightroom Photoshop Fieldwork Remote Sensing QGIS GRASS GIS
Read More

Multi-Agent Research Automation

01/2026
Type: Open-Source Tool

An 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.

Python PydanticAI Mistral Serper Multi-Agent AGPL-3.0
View on GitHub

Ready to transform your geospatial data?

We reply within one business day.