Course

Unit 1

Introduction, Architectures and Design Patterns:Introduction to Agentic AI, Agent Structure and Architecture of multi-agent systems, Autonomous Agents, Human in the Loops Systems, Multi Agent AI Systems, Agentic AI Frameworks, Design Considerations and Best Practices. Modules: Perception, Cognitive, Action, Learning, Collaboration, Security, Design Patterns: Reflection, Tool Use, Planning, ReAct (Reasoning and Acting) and ReWOO (Reasoning with Open Ontology), Multi Agent

Unit 2

Building AI Agents with LangGraph: Introduction to LangGraph, State Management, Trim and Filter Messages, Memory and External Memory, Short and Long Term Memory, Memory Schema, Deployment.

Unit 3

Agentic RAG: Comparing Agentic RAG with Traditional RAG, Agentic RAG Architecture and Components, Adaptive RAG, Variants of Agentic RAG, Applications, Agentic RAG with LlamaIndex.

Unit 4

Agent Development with AutoGen:Autogen Introduction, Roles and Conversations, Conversation Patterns, Developing Autogen-powered Agents, Deployment and Monitoring
Multi Agent Systems with LangGraph: Multi Agent Systems, Workflows, Collaboration, Multi Agent Designs, Workflow with LangGraph

Unit 5

Deploying AI Agent: AI Agent Observability and AgentOps, AI Observability with Langsmith, Monitoring AI Agent Performance, Managing AI Workflows, Implementing AI Experimentation and Observability.

Text Books / References

Text Book:

1. Mastering AI Agents: From Design to Deployment of Autonomous Systems by Elbert Gale January 18, 2025, ISBN-13: 979-8307403822

References:

1. Building AI Agent: Learn Building AI Agents, A Practical Guide for Building Intelligent Agentsby NatenapasFaraksa, March 2, 2024
2. Build Your Own AI Agent: A Hands-On Guide for Beginners by Sujai G Pillai, January 13, 2025, ISBN-13: 979-8306885971

Course Objectives:

To equip students with a comprehensive understanding of agentic AI principles, architectures, and design patterns, enabling them to design, develop, deploy, and monitor autonomous and multi-agent systems using modern frameworks and tools such as LangGraph, Agentic RAG, AutoGen, LlamaIndex, and observability platforms.

Course Outcomes:

• CO1: Apply agentic AI principles to design autonomous and multi-agent systems using appropriate architectures and design patterns.
• CO2: Apply LangGraph for state management, memory integration, and system deployment.
• CO3: Apply Agentic RAG to build adaptive retrieval systems and compare with traditional RAG.
• CO4: Apply Autogen and LangGraph to develop, deploy, and manage collaborative multi-agent workflows.
• CO5: Apply observability tools to monitor, manage, and evaluate AI agent workflows and performance.

CO–PO Mapping: