Fundamentals of Agentic AI

LLMs can generate text.

Agents can think, decide, and act.

In this hands-on workshop, you’ll learn the fundamentals of Agentic AI and build a working AI agent in Python that can reason, use tools, and maintain memory.

This session focuses on first principles, not hype or black-box frameworks. You will clearly understand how agents work internally and how to design them correctly for real-world use cases.

What you will learn

1. What Agentic AI actually is (and what it is not)
2. The Agent Loop: Observe → Think → Decide → Act
3. Core agent components: LLM, tools, memory, control loop
4. How agents use tools to perform actions
5. Why memory is critical for intelligent behavior
6. Common mistakes while building agents

What we will build (live)

1. A goal-driven AI agent from scratch in Python
2. Tool-using behavior (calculator / utilities)
3. Memory-enabled agent that improves responses
4. Mini task/research agent

We will also briefly show how the same ideas translate to popular agent frameworks, so you know how production systems are built.

Who should attend

1. Software Engineers (Backend / Full-stack)
2. Data / ML Engineers
3. Developers exploring AI agents
4. Anyone who wants to move from “using ChatGPT” to “building AI systems”

Pre-requisites

1. Basic Python knowledge
2. Laptop with internet
3. OpenAI / Anthropic API key (any one)

No prior Agent or LangChain experience required.