Vibe Engineering (Coding) Crash Course | Build AI Customer Support Agent with AWS Deployment

In this **Vibe Engineering Crash Course**, I show you how modern AI-powered software is actually built from **LLM fundamentals** to **AI coding agents** to **deploying a real project on AWS**. ## Chapters 0:00 Introduction and Welcome 1:06 Course Roadmap Overview 6:01 Why Learn Vibe Engineering Right Now? 11:05 LLMs Explained: The Intelligence Engine 13:24 The ReAct Loop: How AI Agents Think and Act 17:03 Coding AI Agents vs Plain LLMs 21:44 Prompting Effectively: The Genie Analogy 29:50 Vibe Coding vs Vibe Engineering: Key Differences 32:11 Context Window: The LLM's RAM 38:49 Prompt vs Context: A Critical Distinction 41:55 Agent Skills and Smart Context Management 46:53 Coding Agent Types: CLI, IDE, and Cloud 54:17 GitHub Copilot: Setup, Modes, and Features Tour 1:14:56 Cursor IDE: Features, Settings, and Context Tools 1:34:03 Antigravity IDE by Google: Overview and Tour 1:42:07 Project: Building an AI Customer Support Agent 1:49:48 Planning Phase: Architecture and Mermaid Diagram 2:04:22 Live Build: Agent Generates the Full App 2:10:01 Adding Production Logging and UV Package Manager 2:23:36 Running and Testing the App Locally 2:30:32 Deploying to AWS EC2 Using CLI 2:50:14 Live Demo on AWS + Course Wrap-Up This is not surface-level theory. This is a full practical walkthrough for developers and builders who want to understand how to move from **vibe coding** to **vibe engineering**. Inside this course, you’ll learn: - how **LLMs** work in real product-building workflows - what the **ReAct loop** actually means - the difference between **prompt vs context** - why **context windows** matter so much - what **AI coding agents** are - how **agent skills** help control context and workflows - the types of coding agents in the market: **CLI, IDE, and Cloud** - overview of **GitHub Copilot, Cursor, and Google Antigravity** - how to build a **real AI customer support agent** - how to design the architecture and Mermaid diagram - how to generate th