Doing Prompt Cartography: Practical LLM Workflows for GIS Professionals

Workshop 2:

Doing Prompt Cartography: Practical LLM Workflows for GIS Professionals

Large language models are suddenly everywhere, but most GIS professionals are still trying to answer a very practical question: What can I actually do with this stuff on Monday morning?! 

This workshop introduces prompt cartography as a hands-on, beginner-friendly approach to using LLMs to plan, critique, refine, and communicate better maps. No coding background is required, no AI mysticism will be performed, and no one will be asked to say "agentic system"—ever. (Probably.)

Instead, you'll learn how to turn vague map ideas into structured prompts that produce more useful, more specific, and more defensible results than traditional mapping approaches—in a fraction of the time. 

We will begin with a simple principle: specificity beats cleverness. From there, we will practice writing prompts that clarify purpose, audience, data needs, visual style, map elements, interactivity, and critique. 

The workshop borrows from the prompt cartography workflow outlined in the book Prompt Cartography (2026): data engineering, visual planning, interface planning, critique, and publishing/maintenance as a repeatable professional process rather than one-off AI experimentation. As will be shown, prompt cartography is a specific workflow of data engineering, visual planning, interface/behavior planning, critique, and publishing rather than a magic "data in, map out" trick.

This workshop is designed for GIS professionals and students who want to use LLMs responsibly without losing control of their cartographic judgment. Participants will try short exercises in their own preferred LLM tools using either their own datasets or provided examples. We will compare vague prompts to structured prompts, build small data briefs, describe map styles in usable language, and ask LLMs to critique maps like skeptical-but-helpful cartography colleagues. 

By the end of the session, participants will leave with a practical starter kit: 

• reusable prompt patterns, 
• a simple prompt-to-map workflow, 
• short reflection exercises, and 
• a better sense of where LLMs help, including where they hallucinate, and where human GIS judgment remains non-negotiable. 

The goal is not to replace GIS expertise. The goal is to make that expertise easier to express, delegate, test, and refine through language. Or, to put it another way: GIS software is changing, but you’re still driving the bus.

Requirements for Attendees:

• Registered LLM account(s) of some sort:
  • Presenter will use ChatGPT. Recommended for following along.
    • Sign up for a free account here: www.chatgpt.com
    • Other accounts should work too:
• • • • Google Gemini: www.gemini.google.com
      • Perplexity: www.perplexity.ai
• Code Editor or IDE:
  • Visual Studio Code (free): https://code.visualstudio.com/download 
  • Brackets (free): https://brackets.io/
  • Notepad++ (free): https://notepad-plus-plus.org/ 
• GIS Datasets
  • Acceptable dataset formats include:
    • Shapefiles (zipped)
    • Geopackages
    • Geojson
    • CSV
• Python installed on computer (required to run a local web server to test our interactive maps, no python programming knowledge required).

Recommendations

Pull-the-plug on a streaming service for a month and ante up for a ~$20 a month subscription to ChatGPT, or other capable LLM (see above), before the workshop. If you have a paid account, you won’t run into maximum token issues and you will have the most powerful/capable LLMs at your disposal. 

You can sign up for one-month subscriptions on most LLM sites. It makes a HUGE difference in capabilities and outputs and will make the workshop more rewarding for you. Not required, though.

Real-life Examples

No need for real-life examples. The workshop will demonstrate what is possible and attendees can then use, adapt, and evolve their introductory skills to do infinite things using prompt cartography as a method moving forward. 🙂

Workshop Details:

Length:  3 hours

Cost:  $50 (member) & $100 (non-member)

Date/time:  August 19th, 2026, 9am-12pm

Presenter:  Ian Muehlenhaus, ESRI

Click the Register Now button to participate in this Summer Virtual Workshop!

Can't make the workshop date/time work with your schedule? No worries, a link of the recording will be made available to all those who register for this workshop at a later date.