Prompting techniques
Beyond the basics — chain-of-thought, prompt chaining, role assignment, and format control for predictable, repeatable results.
If you are new to AI, read these in order. The first guide is the 30-minute starter path that links to every other piece. The rest go deeper on the questions everyone hits early: which model to use, what AI gets wrong, what is safe to share at work, and how to prompt.
Chain-of-thought prompting: when reasoning out loud changes the output
Why telling a model to "think step by step" works, and when it does not. Zero-shot CoT vs few-shot CoT, what tasks benefit most, and the cases where it actively slows you down.
Prompt chaining: how to break complex tasks into reliable steps
Why one long prompt fails where a chain of short ones succeeds. How to pass outputs as inputs, where to put the split points, and when chaining is overkill.
Role prompting: how to assign a persona that actually steers behavior
The difference between system-level personas and inline "act as" instructions. When role assignment changes output quality, when it's theater, and the failure modes that make output worse.
Getting structured output: JSON, lists, and formats that hold
Why models drift from requested formats and what actually fixes it. How to get reliable JSON, tables, and structured lists — with and without native structured-output APIs.
Building with prompts
Ready to embed prompts in a product? System prompts, iteration, injection defense, and cross-model porting.