AI Agents 101 – Understanding and Building AI Agents

Description:

This half-day training covers the fundamentals of AI Agents, how they interact with LLMs, and a deeper dive into their architecture. Participants will learn about the technical foundations and get hands-on experience with popular open-source agent frameworks.

Duration: 90 Min

Course Code: BDT404

Learning Objectives:

After this course, you will be able to:

• Understand the fundamental components of an AI Agent
• Design the basic architecture of an AI Agent using popular frameworks like LangChain
• Implement essential components including memory systems and tool integration in a simple agent
• Create and deploy a basic functioning AI Agent using open-source tools

• Basic Python knowledge
• Familiarity with LLMs and APIs
• Basic command line experience

• IT professionals
• Developers
• AI/ML enthusiasts
• Students with basic programming knowledge.

Course Outline:

Module 1: Foundations of AI Agents  

• Agent architecture and design patterns
• Types of AI Agents: Reactive, Deliberative, and Hybrid
• Core components: Perception, Planning, and Execution
• Agent communication protocols

Module 2: Understanding Agent Frameworks

• Overview of popular frameworks
• Deep dive into LangChain
• Introduction to Auto-GPT architecture
• Comparison of different agent approaches

Module 3: Building Blocks of AI Agents

• Natural Language Processing fundamentals
• Prompt engineering for agents
• Memory systems and vector databases
• Tool use and function calling

 Module 4: Hands-on with Basic Agents

• Setting up a development environment
• Creating a simple agent using LangChain
• Implementing basic tasks (retrieval-based answering, automation)
• Adding tool capabilities

Module 5: Advanced Concepts and Case Studies

• Multi-agent systems
• Agent orchestration
• Error handling and reliability
• Best practices and limitations

Hands-on Component - LangChain agent  or smolagent  implementation

Training material provided: Yes (Digital format)