# Columbus Earth

> Columbus Earth is a spatial frontier AI company building the first production Large Geospatial Model (LGM) — an AI that understands the physical world. Products include Columbus Pro (enterprise geospatial intelligence) and Elio (consumer travel and city exploration).

Columbus Earth's mission is to bring a paradigm shift in the relationship between AI and the physical world. The company collects the world's data and builds AI that comprehends it all — from satellite imagery and temporal patterns to public and private geospatial datasets.

## Core Pages
- [Home](https://columbus.earth/): Columbus Earth overview — the Large Geospatial Model, Columbus Pro, and Elio
- [Company](https://columbus.earth/company): Mission, vision, founders, and what Columbus Earth is building
- [Research](https://columbus.earth/research): An interactive thesis on building the Large Geospatial Model

## Products
- [Columbus Pro](https://columbus.earth/products/business): Agentic GIS platform for site selection, due diligence, and location research — for real estate, urban infrastructure, geomarketing, and more
- [Elio](https://columbus.earth/products/consumer): The social super map — AI-powered city and travel exploration for consumers

## Engineering
- [The Philosophy of a Universal Geospatial Model](https://columbus.earth/blog/philosophy-universal-lgm): Our vision for a universally capable model for spatial and geospatial reasoning
- [What is a Large Geospatial Model?](https://columbus.earth/blog/what-is-a-large-geospatial-model): Definition of a Large Geospatial Model, and how it differs from World Models, LLMs and other foundation models
- [Why LLMs Didn't Cut It](https://columbus.earth/blog/why-llms-didnt-cut-it): The problems we encountered when using LLMs to build our products, and examples cases
- [Earth Recipes: How a World Would Think](https://columbus.earth/blog/earth-recipes-world-think): Building accurate geospatial core reasoning with our concept of Earth Recipes
- [Mapping the Unknown: Generative Geospatial Layers](https://columbus.earth/blog/mapping-unknown-gen-layers): How Generative AI would be applied to maps and geospatial data
- [Turning Dormant Data Usable](https://columbus.earth/blog/turning-dormant-data-usable): Our data digestion concept of Data Enzymes to aggregated quality training-data at scale and cheaply
- [The Large Geospatial Model: Our Timeline](https://columbus.earth/blog/lgm-timeline): A taxonomy of different types of Physical AI, where we are, and our aimed release dates.
- [Why We're Building an LGM](https://columbus.earth/blog/why-building-lgm): Three reasons behind the model
- [Geospatial AI's Reference Repository](https://columbus.earth/blog/geospatial-ai-reference-repository): Repository of Geospatial AI papers and works.

## Product Articles
- [A guide to Europe: Cheap but unique cities](https://columbus.earth/blog/elio-europe-guide): I used Elio to find cool hidden cities that are affordable!
- [Chats with Elio: the hidden hill above the city](https://columbus.earth/blog/elio-chat-hidden-hill): A real chat I had with Elio, start to finish!

## Company News


## Optional
- [Blog](https://columbus.earth/blog): All research notes, product releases, and company updates from Columbus Earth
- [Contact](https://columbus.earth/contact): Contact form for demos, investor relations, careers, and general inquiries