Course Content

​

Module 1: Introduction to Agentic AI

• 1.1 What is Agentic AI?

  • Definition and distinction from traditional AI

  • Autonomous vs. agentic behavior

• 1.2 Evolution of Autonomous Agents

  • From chatbots to LLM agents

  • Rise of AI Agents post-GPT

• 1.3 Core Components of Agentic AI

  • Memory, Goals, Tools, Environment, Reasoning loop

• 1.4 Use Cases

  • Task automation, multi-step workflows, RPA, autonomous research

Module 2: Architecture of Agentic AI

• 2.1 Agent Loop & Planning Systems

  • ReAct (Reason + Act)

  • Chain-of-Thought prompting

• 2.2 Tool Usage

  • APIs, plugins, external tools

• 2.3 Memory Systems

  • Short-term vs. long-term memory

  • Vector databases (e.g., FAISS, Pinecone)

• 2.4 Observation, Planning, and Action Cycle

• 2.5 Multi-Agent Systems (MAS)

  • Communication and collaboration among agents

Module 3: Tools & Frameworks

• 3.1 LangChain

  • Agents and tools in LangChain

  • LangGraph and workflows

• 3.2 OpenAI Assistants API

  • Threads, tools, and function calling

• 3.3 AutoGen (Microsoft)

  • Role-based agents and group chats

• 3.4 CrewAI

  • Task delegation and multi-agent coordination

• 3.5 Other Tools

  • HuggingGPT, AgentVerse, BabyAGI, SuperAGI

Module 4: Building Agentic Workflows

• 4.1 Designing an Agent

  • Define goals, constraints, tools

• 4.2 Agent Personas & Role Assignment

  • Specialist vs. generalist agents

• 4.3 Task Decomposition

  • AutoGPT style goal-breaking

• 4.4 Tool Integration

  • File I/O, web search, API calling, database interaction

• 4.5 Chaining & Memory Use

  • Building workflows with persistent context