Google Cloud Next Wrap-Up

To wrap up this year’s Google Cloud Next, The New Stack founder Alex Williams, The New Stack AI editor Frederic Lardinois, and our analyst friend and regular contributor Janakiram MSV sat down to discuss the news from the conference.

Given that the conversation at Google Cloud Next was all about AI agents, it’s no surprise that the discussion here also focused on exactly this topic, with a focus on agent frameworks. As MSV noted, last year was all about third-party vendors like CrewAI, Microsoft with Autogen, Langchain and others launching their frameworks for building models. Now, we’re seeing first-party frameworks from model providers like Google, which announced its Agent Development Kit (ADK) at the event.

“First, it was OpenAI that launched an agent SDK a couple of weeks ago, and now it is the turn of Google to launch something native, which is their agent SDK,” MSV said. “So that is a big deal — and Microsoft is also following suit with [an] OpenAI-based agent SDK. […] So it’s big news and a splash that Google is launching their first party agentic framework.”

It’s still very early, though, and as Lardinois noted, a lot of companies are still figuring out where AI agents can play a role in their workflows.

As MSV noted, Google is now able to tell a comprehensive AI story for developers, from the models to the model runtime, the frameworks and tooling for testing these models and applications, all the up to Vertex AI and other platforms for putting these models into production. It is now building out its story around agents with even more tools for enterprises, including the no-code Agent Space for building agents and the Agent Garden registry.

The discussion also covered how some of the existing cloud native technologies like Kubernetes and the tooling around them, including, for example, OpenTelemetry, fit into this new AI world, as well as how agents will change the developer workflow.