Agentic AI Architectures with Patterns, Frameworks and MCP

Packt via Coursera

Go to class Write review

Overview

Google, IBM & Meta Certificates – 40% Off

One plan covers every Professional Certificate on Coursera.

Unlock All Certificates

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. Unlock the next level of AI mastery by learning how to design and implement agentic AI architectures. This course empowers you to build intelligent multi-agent systems, understand core components like perception, planning, memory, and communication, and leverage frameworks and design patterns to create robust AI solutions. Your journey begins with a deep dive into AI agents, exploring principles, generative AI integration, and real-world use cases. You'll progressively design simple agents, then advance to sophisticated architectures, including orchestration, choreography, and multi-agent systems. Along the way, you'll learn the intricacies of Agentic RAG, communication protocols, and context engineering. The course then guides you through frameworks, patterns, and Model-Context Protocols (MCP), equipping you with hands-on skills to design AI workflows, agent orchestration, and platform integration. Practical projects such as building research assistants, ChatGPT agents, and health agents ensure that theory translates into applied expertise. This course is ideal for AI developers, software architects, and technical professionals aiming to implement agentic AI systems. No prior experience with MCP is required, but familiarity with AI fundamentals and microservices is recommended. Difficulty level: Intermediate to Advanced. By the end of the course, you will be able to design and implement agentic AI architectures, apply design patterns effectively, build multi-agent workflows, utilize Model-Context Protocols, and develop context-aware AI solutions across real-world applications.

Syllabus

Introduction

In this module, we will introduce you to the world of Agentic AI and set the foundation for your learning journey. You will explore the tools and resources that will guide your progress. Finally, you’ll begin applying concepts through a hands-on course project.

What are AI Agents? Key Principles

In this module, we will explore the core principles that define AI agents and their capabilities. You will evaluate when and where AI agents can be applied. Finally, you will gain practical experience by designing your first simple AI agent.

Agent Anatomy - A Deep Dive Into Core Components

In this module, we will dissect the anatomy of AI agents, examining their brains, senses, memory, actions, and communication. You will understand how the Agentic Loop drives agent behavior. Hands-on exercises will enable you to build real-world AI agents.

What is Agentic AI?

In this module, we will introduce Agentic AI and highlight its distinguishing features. You will explore system-level architectures and multi-agent system designs. Hands-on exercises will guide you in designing orchestration and choreography models.

From Traditional to Agentic Workflows

In this module, we will explore the transition from traditional to agentic workflows. You will examine the components, patterns, and frameworks that enable efficient agentic systems. Practical exercises will help you build workflows in real-world scenarios.

The Frameworks Landscape

In this module, we will survey the frameworks that support Agentic AI development. You will explore tools like LangChain, LlamaIndex, Microsoft Agent Framework, and cloud-based platforms. Practical insights will enable you to select and combine frameworks effectively.

The "Why" of Agentic Design Patterns

In this module, we will explore the rationale behind Agentic AI design patterns. You will examine patterns for task management, collaboration, and safety. Hands-on exercises will allow you to apply these patterns in real-world agentic scenarios.

The Limitations of Traditional RAG

In this module, we will examine the constraints of traditional RAG systems. You will explore the improvements offered by Agentic RAG. Practical exercises will illustrate how the Active Researcher loop enhances retrieval performance.

Agent Communication Protocols

In this module, we will address the “Protocol Problem” in Agentic AI. You will explore different communication protocols, including ACP, A2A, AG-UI, and MCP. Architectural insights will help you implement effective communication systems.

Model-Context Protocol (MCP)

In this module, we will introduce the Model-Context Protocol (MCP) and its role in Agentic AI. You will explore MCP architecture, lifecycle, and core concepts. Real-world examples and design patterns will demonstrate its practical application.

Context Engineering

In this module, we will explore the critical role of context in AI systems. You will examine types of context and how they influence agent behavior. Practical strategies and use cases will illustrate effective context engineering.

AI Agent Development Lifecycle

In this module, we will guide you through the AI agent development lifecycle. You will explore evaluation, AgentOps, and enterprise considerations. Hands-on exercises will help you design and manage agentic AI platforms effectively.

Conclusion

In this module, we will review your journey through Agentic AI architectures. You will consolidate your knowledge of design patterns, frameworks, and MCP. Finally, you will reflect on practical applications for future projects.