To demonstrate Looker’s new Conversational Analytics APIs, we’ve added an example custom application to our Conversational Analytics demos and tools repository. This is a complete, executable boilerplate app which you can clone to quickly build custom data agent experiences. The code is TypeScript-first (Next.js + React) and shows how to wire the Looker SDK into your stack to integrate secure, natural language querying in your own applications.
What’s inside: this demo features a custom layout that separates the conversational thread from a dedicated “Insight Workspace” for a side-by-side data exploration experience. The app handles the heavy lifting of real-time chat streaming, multiple agent selection, conversation CRUD operations, and session state management.
We’ve also included an integrated log drawer, making it easy for developers to see exactly which SDK endpoints and payloads are firing under the hood. This gives you a solid foundation to build upon, whether you want to add custom visualizations, embed the interface into an existing portal, or develop advanced multi-agent workflows.
It also includes out-of-the-box support for both Service Account and User OAuth authentication so it can easily adapt to your specific security requirements. Check it out and start building data experiences today.
https://github.com/looker-open-source/ca-demos-and-tools/tree/main/looker-api-ca-demo