About Spatchat

Our Vision

Spatchat.org is an open, extensible platform for conversational spatial analysis, powered by artificial intelligence and large language models (LLMs). It enables researchers, educators, and practitioners to build and share modular chatbot interfaces (i.e., "rooms") for landscape and ecological modeling. Spatchat represents a paradigm shift: it democratizes access to complex geospatial tools by enabling natural language interaction.

Why We Built It

Geospatial tools like GRASS GIS, ArcGIS, and Python are powerful, but often intimidating. They require significant technical expertise, multi-language proficiency, and hundreds of lines of code. Spatchat is a response to that complexity — inspired by real-world teaching experiences and driven by a desire to make spatial analysis more accessible, especially for students and early-career professionals.

What It Can Do

• Chat-based access to spatial tools — ask questions like “Compare class 1 and 2” or “Calculate edge density.”
• LLMs translate natural language into code, run models, and return clear results.
• Support for Python, R, Java, REST APIs — all unified in a simple interface.
• Cloud-native: no local installation required.
• Designed to support modular, community-contributed tools.

Architecture Overview

Each Spatchat room combines a conversational UI, an intent parser powered by LLMs, and a backend spatial engine. Users interact through a lightweight chat interface that can be embedded in websites, notebooks, or apps. Under the hood, python libraries like PyLandStats or models like MaxEnt run based on user prompts — and return structured outputs, charts, and visualizations.

Looking Ahead

Spatchat aims to become a central hub for spatial modeling bots. Our goal is to reduce friction, lower barriers to entry, and foster a community of open collaboration. We're especially excited about the potential for Spatchat to reach younger learners — enabling high school students to engage with landscape ecology through natural language long before they’d normally touch GIS or programming.

Join Us

We invite collaborators — researchers, developers, teachers, and students — to help shape the future of spatial intelligence. Contribute a room, suggest an idea, or simply explore what's possible when geospatial reasoning becomes conversational.