Overview

This course features Coursera Coach! A smarter way to learn with interactive, real-time conversations that help you test your knowledge, challenge assumptions, and deepen your understanding as you progress through the course. In this comprehensive course, you will delve into the world of AI agentic workflows, exploring the foundational design patterns that make intelligent systems tick. You'll gain hands-on experience in building and integrating these design patterns, starting with the Reflection pattern, followed by Tool Use, Planning (ReAct), and Multi-agent collaboration. As you progress, you will learn how to implement these patterns in real-world scenarios, enhancing your ability to design AI systems that are both efficient and intelligent. The course is structured into multiple modules, each focusing on one design pattern. The first few lessons set the stage by covering essential prerequisites, such as development environment setup and an in-depth dive into AutoGen, a tool crucial for AI agentic workflows. Then, each design pattern is introduced through theory, followed by hands-on exercises to cement your understanding. Whether you're creating simple workflows or complex multi-agent systems, the course guides you step-by-step through the process. This course is perfect for anyone with a basic understanding of AI looking to deepen their knowledge and practical skills. You’ll learn how to apply cutting-edge techniques to solve real-world challenges in AI agentic systems. The course is beginner-friendly, but a background in basic AI principles and programming would be beneficial. By the end of the course, you will be able to confidently apply the four key AI agentic design patterns to build sophisticated workflows, integrate multi-agent systems, and optimize decision-making in intelligent systems. You will also understand how to leverage Ollama Models locally to reduce costs in AI workflows.

Syllabus

Introduction

In this module, we will introduce you to the course’s structure, the target audience, and the key objectives that will guide your learning. You will also set up the development environment to ensure you're prepared to dive into hands-on exercises. Finally, we’ll provide a crash course on AutoGen, offering a solid foundation for building AI agentic workflows.

AI Agentic Design Patterns - Deep Dive

In this module, we’ll introduce the concept of design patterns and how they’re applied to AI agentic workflows. We’ll explore the key AI agentic design patterns and examine how they help structure and streamline intelligent systems. By the end, you’ll understand their critical role in AI development.

The Reflection Pattern - Hands-on

In this module, we’ll dive into the Reflection pattern and its application within AI workflows. You’ll gain hands-on experience by building out the Reflection pattern and integrating it with orchestration elements to create a more robust AI system. You’ll also learn to use Ollama Models locally as a cost-free alternative to the OpenAI API, providing you with an efficient, practical tool for AI development.

The Tool Use Pattern

In this module, we will introduce the Tool Use pattern and demonstrate how it’s implemented in AI agentic workflows. Through hands-on exercises, you’ll build a trip planner workflow, applying this pattern to real-world tasks. By the end, you’ll understand how the Tool Use pattern can streamline AI workflows and enhance functionality.

The Planning - ReAct - Pattern

In this module, we’ll introduce the Planning pattern and ReAct technique, which are essential for decision-making and strategic planning within AI systems. You’ll gain hands-on experience by applying these patterns in practical workflow scenarios. By the end, you’ll have a clear understanding of how these tools support intelligent decision-making in complex tasks.

Multi-agent (Collaboration) Pattern

In this module, we’ll dive into the Multi-agent Collaboration pattern, which facilitates cooperation between multiple AI agents to achieve complex goals. Through hands-on exercises, you’ll explore how these agents work together, gaining valuable experience in developing multi-agent systems. By the end of this section, you’ll be able to design and implement workflows with multiple collaborating agents.

Wrap-up

In this final module, we will summarize your key learnings and celebrate your accomplishments. We’ll discuss next steps and available resources to help you continue growing your knowledge and skills in AI agentic workflows. By the end, you’ll have a clear roadmap for advancing your AI expertise.