Building Agents with Vertex AI

51mGeneral2024-09-30

Authors

Morten Rand-Hendriksen

Senior Staff Instructor, Speaker, Web Designer, and Software Developer

Course details

AI agents can serve as natural language interfaces to complex data systems by assisting users in finding data, contextualizing data, and performing tasks. In this course, you’ll learn how to use the Vertex AI Agent Builder—a powerful low-code environment—to build networks of AI agents that work together to access data and solve complex tasks through a chat interface. Through practical examples, you’ll explore how to build agents that query documents, connect to APIs, and use custom Python tools to query, retrieve, and synthesize data from an external website.

Learning objectives
Build an AI agent using Vertex AI.
Explore how to ground the AI agent in documents using automatic RAG.
Extend an AI agent by connecting it to an API using a custom OpenAPI schema.
Build a custom Python tool for an AI agent to query and synthesize live web content.

Skills covered

Vertex AIGenerative AIGoogleArtificial Intelligence (AI)One-Off

Concepts

0. Introduction

01 - A low-code tool to build advanced AI agents
02 - Getting a Google Cloud account to follow along

1. Building a Basic Vertex AI Agent

03 - Navigating the Google Cloud Ecosystem
04 - Creating a new Google Cloud project
05 - Accessing the Vertex Agent Builder
06 - Creating a new Vertex AI Agent
07 - What the octopus can teach us about Vertex AI Agents
08 - Building a basic agent
09 - Fine-tuning the agent with examples

2. Extend Vertex AI Agents with Tools and Sub-Agents

10 - Using tools for no-code RAG, part 1
11 - Using tools for no-code RAG, part 2
12 - Using tools for external API lookup, part 1
13 - Using tools for external API lookup, part 2
14 - Using tools for external API lookup, part 3
15 - Making agents work together
16 - Previewing course documentation

3. Conclusion

17 - Vertex AI Agents and where to go from here