Back to Applied AI Engineering
💬
Phase 2
Prompt Engineering & Structured Outputs
Zero-shot, few-shot, and chain-of-thought patterns; how to make a model's output a reliable contract instead of a guess; and why a system prompt has to be engineered defensively, not just written politely.
1
Standard Prompt Patterns: Zero-Shot, Few-Shot & Chain-of-Thought
System, user, and assistant roles, and the three prompting patterns that get a model reasoning about the right thing before you ever touch output format.
prompt-engineeringfew-shotchain-of-thought
2
Structured Outputs & Function Calling
Why 'respond with only the category name' fails silently, and how schema-constrained decoding and tool calling turn a model's output into a real contract.
structured-outputsfunction-callingtool-calling
3
Prompting as Interface Design: Defending Against Injection
Why the system prompt isn't a hard boundary, how indirect prompt injection works, and the defense-in-depth mitigations that reduce — not eliminate — the risk.
prompt-injectionsecuritysystem-prompt