What is prompt engineering?
Prompt engineering is the practice of designing and testing prompts so an AI model reliably gives the output you need. It uses techniques like giving the model a role, examples, a clear format, step-by-step instructions and constraints, then checking results and refining.
How does prompt engineering work?
You write a prompt, look at several outputs, notice what goes wrong, and adjust: add an example, specify the format, or split the task into steps. For apps, prompts are tested against many inputs, not just one.
Why does prompt engineering matter?
Better prompts mean better answers without paying for a bigger model. It's also the first step to building reliable AI features.
An example
Example: A support team adds three example replies to its prompt, and the AI's answers start matching the company's tone.
Related terms
Learn it properly
Prompt engineering is covered in Use AI Well, with short lessons, practice questions and instant feedback.