Skip to content
Courses

Prompt

A prompt is what you write to an AI model to ask it for something: the question or instruction, plus everything you give it to work with, like data, examples or the format you want back.

Select any words here to ask your Maestro about them.

You write to an AI: "Analyze my sales and tell me how we're doing." Of these five decisions, how many will it have to guess?

Each view adds to the same prompt. Filled dot: you decide; ring: the AI guesses.

What it's for: the prompt is your only lever on what an AI model does. An LLM knows nothing about your work beyond what you write and hand it: whatever isn't in the prompt (or in the files you pass along), it guesses.

A good prompt for data work has five parts: the task in one sentence, the data (and what a row is), what counts and what doesn't, the shape of the answer, and how you'll check it. It doesn't need to be long; it needs to leave no important gaps. Showing the AI an example of what you want beats describing it.

Always ask it to show its work: how many rows it read, what it dropped, which formula it used. That makes its answer checkable, and a hallucination shows up before it reaches your report.

You practice it in How do you ask an AI for a number you can check?.

How it's used

A prompt you can check

For the October report

  1. Calculate October's average ticket by store.The task, in one sentence.
  2. Use sales_october.csv: one row per item sold; a ticket spans several rows.The data, and what a row is. Without this, it averages items, not tickets.
  3. Count paid tickets only; leave out cancellations.What's in and what's out.
  4. Give me a table: store, tickets, total and average.The shape of the answer.
  5. marked: Tell me how many rows you read, how many you dropped and why.How you'll check it. The part almost nobody writes.
Each line takes away a decision the AI would otherwise make quietly on its own. The last one gives you a way to check its number.

When to look closer

Is a longer prompt a better one?

A six-line prompt

  1. Hi, hope you're doing well!
  2. You are an expert data analyst with 20 years of experience.
  3. My boss needs this for Monday's meeting and it's really important.
  4. Use sales_october.csv and count paid tickets only.
  5. Please be very detailed and accurate.
  6. Give me the average ticket by store in a table.

How many of these six lines change what the AI will calculate?

Ask your Maestro about Prompt

It answers from the free course page closest to this word, the way it answers while you study.

Example

You: What does "Prompt" mean, in plain words?

Your Maestro: I'll explain it with a case of yours: it sticks better than a definition. What did you last ask an AI, and what did it have to guess?

Asking about: Answers from the page "How do I ask Claude, ChatGPT or Copilot for a dashboard? Write a brief"

Only this browser keeps what you ask here. How it works

Learn more

Related terms

  • LLM (large language model)

    An LLM is a program trained on huge amounts of text to continue a piece of text one chunk at a time: it works out which chunk is likeliest to come next and writes it. ChatGPT, Claude and Gemini all run on an LLM.

  • Context window

    Everything a language model has in front of it at a given moment: your instruction, the files it has read and the results of its tools. What isn't in the context, the model can't use.

  • AI hallucination

    A hallucination is an answer from an AI model that sounds confident and well written but is false or comes from no source at all: a number, a fact or a quote nobody ever said.

  • Agent instructions file

    A text file in your project that the agent's harness loads at the start of every session, so the agent knows from the start what it always needs to know: rules, facts about the data and how to check its work.

The whole glossary