Prompt engineering

01why it matters for a business

The same model can give mediocre or excellent results depending on how it is briefed. For a business, prompt quality shows up in two places. For employees using AI assistants day to day, it is a skill like writing a good brief for a colleague: context, goal, constraints, and what a good result looks like. For AI features built into software, prompts are part of the product and deserve the same discipline as code: version control, review, and testing.

The term can make the work sound like a bag of tricks. In practice it is clear communication plus measurement. Modern models need fewer tricks than early ones did, but they still cannot read your mind about audience, format, or what to do when information is missing.

02what it looks like in practice

A sales team uses AI to draft follow-up emails after discovery calls and finds the drafts generic. The revised prompt includes the call notes, the prospect's industry and role, the two or three points that matter to that buyer, the company's positioning in a short paragraph, an example of a strong past email, and an instruction to ask a clarifying question rather than invent details that are not in the notes. Drafts go from rewritten every time to lightly edited, and the prompt lives in a shared template so the whole team benefits.

03common mistakes

• Changing prompts in production without testing them against a set of real cases.
• Leaving out context the model cannot know, such as audience, purpose, and constraints.
• Long lists of rules with no example of a good result.
• Treating prompting as a substitute for data. If the facts are not in the prompt or retrieved, the model cannot use them.

04related terms

• System promptA system prompt is the standing set of instructions an application gives a language model before any user input, defining its role, rules, tone, available tools, and limits.
• AI evaluations (evals)Evals are structured tests that measure how well an AI system performs on a defined task, using a set of real or realistic inputs with known good outcomes.
• Fine-tuningFine-tuning is the process of further training an existing AI model on a curated set of your own examples so it consistently produces a particular style, format, or behavior.
• Large language model (LLM)A large language model (LLM) is an AI model trained on very large amounts of text to predict the next piece of text, which lets it write, summarize, translate, classify, extract information, and reason through problems in everyday language.

05where insomnia club fits