Track 07: Prompt Engineering

The interface layer between "a model that predicts tokens" and "a system that reliably does what you need." Covers the techniques that measurably work (few-shot, chain-of-thought, structured output), how to organize and version prompts like code, and — critically for anything that touches untrusted input — prompt injection and how to defend against it.

Modules

1. How Prompts Actually Work
2. Anatomy of a Good Prompt
3. Zero-Shot Prompting
4. Few-Shot Prompting
5. Chain-of-Thought
6. Self-Consistency and Voting
7. System Prompts and Personas
8. Prompt Templates and Variables
9. Structured Output Prompting
10. Delimiters and Context Organization
11. Prompt Chaining
12. Reflection and Self-Critique
13. Prompt Injection Attacks
14. Defending Against Injection
15. Prompt Versioning and Management
16. Prompt Evaluation and Iteration
17. Capstone Project