11-17, 09:00–12:00 (Pacific/Auckland), WF503
Combine deck.gl visualization, data processing, and AI APIs to create intelligent mapping applications. Build a real geospatial AI app from scratch, processing spatial data and generating insights through modern web technologies.
This hands-on workshop guides developers through building a complete geospatial AI application using modern open source tools. Participants will create a functional application that processes spatial data, generates AI-powered insights, and presents results through interactive visualizations - all running in the browser with no backend required.
Technology Stack
- Visualization: deck.gl for high-performance geospatial rendering
- App development: Vercel AI SDK
- AI Integration: AI APIs for intelligent data analysis
- Cloud Analysis: Using MCP to access data and run cloud analysis from CARTO platform
- Development: React with TypeScript
Workshop Structure
Hour 1: Foundation
- Project setup with React, TypeScript
- Simple LLM tool integration
Hour 2: AI & data Integration
- Loading and processing geospatial datasets (GeoJSON, CSV)
- Controlling UI via AI
Hour 3: Visualization and Integration
- Creating interactive maps with deck.gl
- Connecting CARTO maps & external tools
- Postprocessing
Requirements
- JavaScript/TypeScript and React experience
- Laptop with Node.js and modern browser
- Anthropic API key
Participants will leave with a complete, functional geospatial AI application and reusable code patterns for building similar applications.
Master in Physics by the University of Oxford. Primary maintainer of deck.gl, the leading open-source framework for high-performance web geospatial visualization. Creator of A5, an innovative pentagonal spatial indexing system with truly equal area cells that offers superior accuracy and lower distortion than traditional hexagonal approaches.