Coursera – Building AI Agents and Agentic Workflows Specialization

Build Autonomous Applications with AI Agents. Master Agentic AI with multi-agent frameworks such as LangGraph, CrewAI, BeeAI, and AG2 (AutoGen)

What you’ll learn
+ Develop agentic AI applications that integrate tools, support reasoning, and improve performance through reflection
+ Design AI agents with LangGraph using agentic AI design principles such as memory, iteration, RAG, and workflow patterns
+ Orchestrate agentic multi-agent systems with CrewAI for collaboration, coordination, and workflows
+ Build conversation-driven agentic AI assistants with BeeAI and AG2; compare AI frameworks for real-world use case applicability

Skills you’ll gain
• AI Orchestration
• Context Management
• Software Design Patterns
• Agentic systems
• Retrieval-Augmented Generation
• LLM Application
• Artificial Intelligence and Machine Learning (AI/ML)
• Agentic Workflows
• LangGraph
• AI Workflows
• Responsible AI
• Application Design
• Large Language Modeling
• LangChain
• Vector Databases
• Prompt Engineering
• Generative AI Agents
• Generative AI
• Tool Calling
• Software Development

Ready to build the next generation of AI applications? This specialization from IBM experts equips you with the skills to develop agentic AI systems using modern frameworks and workflow patterns.

You’ll start with LangGraph, creating agents that support memory, iteration, conditional logic, and retrieval-augmented generation (Agentic RAG).

Next, you’ll explore self-improving agents that use reflection and reasoning, and design multi-agent systems that collaborate through orchestration. With CrewAI, you’ll learn to structure agents, tasks, and tools into modular workflows that solve real-world problems.

Finally, you’ll expand your toolkit with frameworks like AG2 (AutoGen) and BeeAI, applying them to cases such as question answering, summarization, and conversation-driven applications. You’ll also study design patterns like sequential and routing to make systems scalable and reliable.

You will apply the concepts you’ve learned using hands-on labs to build Agentic systems powered by LLMs such as OpenAI GPT, Meta Llama, and IBM Granite.

By the end of this program, you’ll be able to compare frameworks, apply AI design patterns, implement orchestration, and build AI systems that support multi-agent collaboration and advanced workflows. These are the sought-after skills that employers look for in Software Developers, Machine Learning Engineers, Data Scientists, and GenAI Engineers.

Applied Learning Project

In this specialization, you’ll complete hands-on labs that guide you in designing agent workflows, configuring multi-agent systems, & integrating tools into structured applications. Access to cloud-based lab environments are provided to you at no extra cost. By the end, you’ll have built working prototypes with LangGraph, CrewAI, BeeAI, & AG2 (AutoGen) to tackle real-world challenges in collaboration & conversation-driven tasks.

Some examples of the labs & projects are

Build an AI Math Assistant with LangChain

AI Powered Data Analysis with LCEL

Build Interactive LLM Agents with Tools

Building a Reflection Agent with LangGraph

Building a Reflexion Agent with External Knowledge Integration

ReAct: Build Agents that Reason Before Acting

DocChat: Build a Multi-Agent RAG System

Implement CrewBase and Structuring a Crew

Create a Structured Meal & Grocery Planner with CrewAI

Building Agentic AI systems with BeeAI

Build a Multi-Agent Chatbot with AG2 (AutoGen) for Healthcare