New Conversational Analytics API + ADK Demo

Looker News Announcements
,
1

We’ve added a new example to our demos and tools repository that pairs the Conversational Analytics (CA) API with the Agent Development Kit (ADK). Unlike the CA API Golden Demo, which focuses on exploring the CA API’s chat features and sample datasets, this demo shows how you can integrate the CA API directly into a multi agent system built with ADK:

ca_api_adk_s

In this demo, you’ll see how to parse and respond to conversations programmatically, making it a perfect starting point for developers building custom conversational agents, analytics pipelines, or AI-driven assistants.

You can run and deploy the demo in a variety of ways - locally via a lightweight server, fully managed on Cloud Run or GKE, or on Vertex AI Agent Engine to use with Agentspace. Whether you’re testing locally or gearing up for production, this example makes it easy to experiment with both the CA API and ADK in one place.

Check it out on GitHub here: Conversational Analytics API with ADK

2

What would be the benefits of using ADK against the API or SDK directly?

3

Generally, I’d recommend using ADK if you want an easier way to build multi agent workflows with built-in orchestration, tools, memory, observability, etc. Use the CA API/SDK directly if you just need a single conversational data agent and are happy handling the wiring up yourself. The latter will ultimately give you more control but will take more time to develop.