All termsGlossary

What is prompt engineering?

Prompt engineering is the work of writing and refining the instructions given to an AI model so that it responds in the way a task needs. It changes what the model is told, not the model itself.

Why it matters

The same model can give a helpful answer or a poor one depending on how it is instructed. In customer service, the prompt sets the agent’s tone, the topics it covers, the steps it follows and the point at which it passes a conversation to a person. Because prompts are plain text, a team can improve behaviour in hours without retraining anything.

How it works

A writer drafts instructions that describe the agent’s role, the rules it must follow, the information it may use and the format of its replies. Sample conversations are often included to show what a good answer looks like. The team then runs test questions, reads the results and edits the wording. Each version is saved, so a change that makes answers worse can be undone.

In practice

A parcel company’s chat agent keeps writing long replies to simple tracking questions. The team adds an instruction to answer in two sentences and to give the delivery date first. They rerun fifty saved customer questions and compare the new replies with the old ones. The answers are shorter and still correct, so the revised prompt goes live that afternoon.

Get Started

Let’s build AI agents, together.

Talk with us to see the top three use cases for AI agents at your business, and how quickly you could go live.

Live in 7 days · Works inside the tools your team already uses · Hands off with full context