AI inception: Using AI to build a Grafana plugin that uses AI

One of Grafana’s greatest strengths as a data visualization tool is that it’s highly composable and extensible. The Grafana plugin platform plays a key role in this, enabling developers to build data sources and application plugins that let users query and visualize data in new, meaningful ways.

AI is now rapidly transforming the Grafana plugin ecosystem, both in how plugins are built and what users can achieve with them.

In this hands-on lab, you will experience firsthand how large language models (LLMs) can simplify plugin development and enrich the user experience. You’ll start by learning about the various types of Grafana plugins, including data source plugins and application plugins, and their use cases. Then, you’ll use Claude LLMs to build a data source plugin, and create an application plugin that uses AI to visualize data from that source. 

By the end of three hours, you will: 

• Understand the various types of Grafana plugins and how they’re used to extend Grafana 
• Leverage Claude LLMs to build a Grafana data source 
• Use prompts to create a configuration editor and query editors  
• Enable end users to interact with AI/LLMs directly from your newly built application plugin

Requirements:

• Laptop with WiFi access
• Linux or MacOS
• Free GitHub user account (no need to be part of any org)
• GitHub Codespaces template (will be provided by workshop leaders)
• A review of Grafana developer plugin tools and key concepts documentation 
• Some git knowledge

This session is tailored for developers and anyone curious about how to build a plugin mostly using prompting.