Prompt engineering: Prompt engineering is the practice of writing and refining instructions to get reliable, high-quality output from an AI model. It covers giving context, showing examples, specifying format, and iterating on what did not work.
The name sounds technical, but the work is closer to writing a good brief than to programming. You decide what the model needs to know, state it clearly, and check the result against what you meant. Then you adjust the prompt, not the output, so the next hundred runs come out right.
The techniques with the biggest payoff are simple. Give the model a role and an audience. Paste real context instead of describing it. Show one or two examples of the output you want (few-shot prompting). Ask for a specific format. Tell it to ask clarifying questions before answering if the request is ambiguous. Ask it to check its own work against a list of criteria.
Prompt engineering matters most when a prompt will be reused: a template your team runs daily, a system prompt inside a custom assistant, or a step in an automation. For one-off questions, a plain conversational request and a quick follow-up is usually enough.
Example at work
A recruiting team's first prompt for screening summaries produced vague write-ups. The recruiter rewrote it to include the job's five must-have criteria, two examples of a good summary, and an instruction to rate each criterion as met, partially met, or not evidenced with a quote from the resume. The summaries became consistent enough to compare across candidates.
Why it matters
You do not need to memorize frameworks, but you do need a habit of specifying, testing, and refining. Teams that treat prompts as reusable assets, saved and improved over time, get compounding returns; teams that retype vague requests each time never get past mediocre results.
Related terms
- PromptA prompt is the text (and sometimes files or images) you give an AI model to tell it what you want. A good prompt states the role the AI should play, the task, the relevant context, the output format, and any constraints.
- Few-shot promptingFew-shot prompting is including a handful of examples of the input and output you want inside your prompt so the AI model can copy the pattern. It is the fastest way to get consistent formatting, tone, and judgment calls.
- Chain-of-thought promptingChain-of-thought prompting asks an AI model to work through a problem step by step before giving its final answer. Showing its reasoning tends to improve accuracy on math, logic, and multi-step tasks and makes mistakes easier to spot.
- System promptA system prompt is a set of instructions given to an AI model before a conversation starts that defines its role, tone, rules, and boundaries. Vendors write one for their products; you can add your own in custom instructions, projects, and custom assistants.
- Zero-shot promptingZero-shot prompting is asking an AI model to do a task with instructions only, without showing it any examples of the output you want. It works well for common tasks the model has seen many times, such as summarizing or translating.