Vertex AI Agent Builder

We will start developing our RAG application step by step, looking at different components. First, let’s start with Vertex AI Agent Builder, a managed Google Cloud service that allows you to build your RAG application based on your data in a few clicks [8], and then we’ll see how we can replace its components one by one with Do-It-Yourself ones.

Creating a data store and indexing documents

When you start developing an RAG application with Vertex AI Agent Builder, your first step should be putting your documents together into a data store. It will automatically index them and prepare them to be used by your RAG application. We’ll cover more technical details about ingesting data into Vertex AI Agent Builder in the next chapter, but for now, let’s start with a bunch of PDFs and other supported formats that you have put together.

First, we need to make our documents accessible for Vertex AI Agent Builder. We do this by creating...