What is prompt engineering in simple words?

Prompt engineering is writing clear instructions for an AI so it gives the answer you need. It means stating the task, adding the context it cannot guess, showing an example and saying what format you want, then improving the prompt by testing it.

Is prompt engineering still useful in 2026?

Yes. Models have become much better at guessing intent, but they still cannot know your situation, your audience or the format your work needs. Clear briefs save time with any model, and teams that build AI apps still write and test prompts every day.

Do I need to be polite to the AI?

Being polite does no harm, but it is not what makes answers good. Clear context, a specific task and a stated format matter far more than “please” and “thank you”.

Is there one perfect prompt template?

No. Role, context, task, example, format and tone is a checklist, not a formula. Use the parts the job needs; a quick fact question may need only the task, while a cover letter needs nearly all of them.

Module 3 · Talking to AI: prompting · Lesson 13 of 32

A prompt is a brief to a new intern

11 minFree GenAI course · BeginnerLesson 3.1By Saket Bhatnagar

Researched and drafted with AI, reviewed and edited by Saket Bhatnagar.

In short

A prompt is a brief. The clearer the brief, the better the work you get back.
Strong prompts give a role, the context, the task, an example, the format and the tone.
The AI cannot read your mind: anything you leave out, it fills with a generic guess.

Soon

Video lesson · coming soon

How to write a good prompt: brief AI like a new intern

Lala, your AI mentor, will walk you through this lesson in a short animated video. Until then, everything is here to read and try.

01 The intern who made the wrong poster
02 AI continues exactly what you give it
03 Meera rewrites the brief
04 Role, context, task, example, format
05 Watch a leave application improve
06 Read, fix one thing, try again
07 Try the prompt builder yourself

Meera runs a small design team in Pune. On her new intern Kabir’s first day, she says, “Make a poster for the Diwali party.” Kabir is bright and fast. By lunch he shows her a gorgeous poster, sized for Instagram, all in English, with last year’s date on it.

Kabir wasn’t careless. The brief was. Every AI chat app is a Kabir: smart, quick, eager to help, and completely new to your situation. This lesson is about writing the brief that Meera should have written.

Why does an AI need a brief at all?

Remember the one trick behind chat AI: the model keeps predicting the next likely token. Your message is the beginning of the text it continues. So everything it writes is a likely continuation of exactly what you gave it, and nothing more.

A short, vague message has thousands of likely continuations. With no clues to narrow them down, the model drifts to the safest middle one. That is why a bare “write a poem about rain” comes back sounding like every rain poem ever written.

The word “engineering” sounds grand, but there are no magic words here. Most of it is clear writing, the same skill a good manager uses to brief a person. If a smart stranger could do the job well from your message, the AI usually can too.

Think of it like this

Meera rewrites the brief

The next morning Meera tries again. “You’re designing for our office notice board. About sixty people, most of them prefer Hindi. The party is this Friday at 6 pm on the terrace. Here’s last year’s poster so you can see our style. I need it A3, print-ready, Hindi on top and English below. Fun, but not loud.” Kabir nails it in one go. He did not get smarter overnight. He finally had what he needed.

In real lifeIn AI

“You’re designing for our office notice board”: In AI: Role: who the AI should act as
Sixty people, Hindi first, Friday 6 pm, terrace: In AI: Context: the facts only you know
“Make a poster for the party”: In AI: Task: the one job to do
Showing last year’s poster: In AI: Example: what good looks like
A3, print-ready, fun but not loud: In AI: Format and tone: the shape and feel of the answer

The parts of a good prompt

You don’t need every part every time. A quick question like “what is the capital of Assam?” needs only the task. But the moment the answer depends on your situation or has to fit somewhere, the other parts start earning their place.

Six parts of a prompt, using a leave application as the example
Part	The question it answers	Example line
Role	Who should the AI be?	You are an assistant who writes polite emails for Indian offices.
Context	What is my situation?	I need leave from 12 to 14 November for my cousin’s wedding in Jaipur.
Task	What exactly should it do?	Write a leave application email to my manager.
Example	What does good look like?	I usually open with: “Hi Ms Rao, hope your week is going well.”
Format	What shape should the answer take?	A subject line, then under 120 words, signed “Sneha”.
Tone	How should it sound?	Warm and respectful, not stiff.

How the parts become one answer

Drag to explore

Each part narrows down what the model is likely to write, so the answer lands closer to what you meant.

Of all the parts, context is the one people skip most and miss most. The model has read a huge library, but it has never met your manager, seen your syllabus or tasted your mother’s dal. Only you can hand it those facts.

Try it: build a leave application prompt

Let’s feel the difference instead of reading about it. The task below stays the same the whole time. Only the brief around it changes.

Try it yourselfPrompt builder: one task, five upgradesStart with the bare task, then switch on Role, Context, Example, Format and Tone one at a time. Watch the assembled prompt, the answer and the clarity meter change.

With only the task, the meter reads “Vague” and the answer is full of placeholders like [dates] and [Your name]. Switch on Context and the biggest change happens. The real dates (14 to 16 March) appear, with the reason (a sister’s wedding in Jaipur), the manager’s name and who covers the work.

The other parts polish the answer. Role and Tone change the wording. Example or Format turns one long paragraph into a proper email with a subject line. With all five on, the meter reads “Perfect brief”.

Here is a full version you can copy into any AI chat app. Swap in your own details, and keep the labels.

text

Role: You are an assistant who writes polite, professional emails for Indian offices.

Context: I am a software tester at a mid-size company in Bengaluru. My manager is Ms Rao. I need leave from 12 to 14 November for my cousin's wedding in Jaipur. My teammate Farhan will cover my test runs while I am away.

Task: Write a leave application email to my manager.

Example of my usual opening: "Hi Ms Rao, hope your week is going well."

Format: A subject line, then the email in under 120 words. Sign off as Sneha.

Tone: Warm and respectful, not stiff.

Prompting is a conversation, not a spell

Even Meera’s second brief might need a tweak. Maybe the Hindi line is too formal, or the date is too small. She does not start over. She says, “Lovely. Make the date twice as big.” Good prompting works the same way.

Write a first brief
Use the parts that matter for this job. Don’t aim for perfect.
Read the answer like a manager
Ask: what is wrong, missing or off in tone?
Find the missing part
A wrong assumption means missing context. A wall of text means missing format.
Add or change one thing
Reply in the same chat, or edit the original prompt and run it again.
Save what works
Keep good prompts in a note so you can reuse them next time.

Apps use this idea behind the scenes too. When a food delivery app’s help bot greets you politely and only talks about orders, that is because the app gave the model a brief before you typed anything.

Check yourself

Riya types “Write a short bio for me” and gets a bland paragraph about “a passionate individual”. What should she add first?

Pick an answer to see why it is right or wrong.

Check yourself

Which of these lines is the format part of a prompt?

Pick an answer to see why it is right or wrong.

Remember these three

A prompt is a brief to a smart newcomer: it can only use what you tell it.
Role, context, task, example, format and tone each narrow the answer towards what you meant.
Treat prompting as a conversation: read the answer, find the missing part, add it, repeat.

Questions people ask

What is prompt engineering in simple words?: Prompt engineering is writing clear instructions for an AI so it gives the answer you need. It means stating the task, adding the context it cannot guess, showing an example and saying what format you want, then improving the prompt by testing it.
Is prompt engineering still useful in 2026?: Yes. Models have become much better at guessing intent, but they still cannot know your situation, your audience or the format your work needs. Clear briefs save time with any model, and teams that build AI apps still write and test prompts every day.
Do I need to be polite to the AI?: Being polite does no harm, but it is not what makes answers good. Clear context, a specific task and a stated format matter far more than “please” and “thank you”.
Is there one perfect prompt template?: No. Role, context, task, example, format and tone is a checklist, not a formula. Use the parts the job needs; a quick fact question may need only the task, while a cover letter needs nearly all of them.