7-day AI challenge for work

The best way to learn to use AI at work is to use it every day on real tasks. Reading about it helps very little. This challenge gives you one task a day for a week, each with a ready-made prompt to adapt.

You do not need to get everything perfect. The goal is to build the habit and find out which tasks AI saves you the most time on.

Before you start, one rule for the whole week: do not paste personal data, passwords or confidential information. When needed, swap names and numbers for examples. See how to use AI at work safely.

Day 1: organize your week

These are my tasks for this week: [list the tasks, one per line].
Sort them into three groups: urgent, important and can wait.
For each urgent task, suggest a first step I can do in 15 minutes or less.

What to watch: the AI does not know your deadlines unless you tell it. If the order looks off, say which tasks have fixed dates and ask it to reorganize.

Day 2: write an email you have been putting off

Write an email to [who] about [topic].
Context: [what happened and why you are writing].
Friendly and direct tone, under 120 words, with a subject line.
End with a clear ask: [what you expect from them].

What to watch: compare it with the email you would write. Change what does not sound like you and save the prompt for next time.

Day 3: summarize a long document

Summarize the text below for someone who has no time to read it all.
Include: the goal in one sentence, the 5 main points as bullets
and the actions someone needs to take, with owner and deadline if given.
If something is unclear in the text, mark it as "to be confirmed".

Text:
[paste here]

What to watch: check that the numbers and dates in the summary match the original. Summaries can get details wrong.

Day 4: create some content

Write a LinkedIn post about [topic] for [audience].
Open with a sentence that grabs attention, without clichés.
Under 150 words, in short paragraphs, ending with a question for the reader.
Give me 3 versions with different angles.

What to watch: ask for changes in the conversation, such as “version 2 is the best, make it shorter”. Adjusting is usually faster than asking for everything again.

Day 5: analyze data and make a decision

This is data about [what the data is]: [paste a small table or describe it].
What are the 3 most important patterns? Explain each in one sentence.
Then suggest a practical decision based on these patterns and tell me which
figure I should check to confirm the decision is right.

What to watch: the AI can point out patterns that do not exist. Check the conclusions against the data itself before acting.

Day 6: solve a problem

I have this problem at work: [describe the situation, without names].
I have already tried: [what was done].
Suggest 5 different solutions, from simplest to most involved.
For each, give the first step and one risk.

What to watch: the first ideas tend to be obvious. Ask for “5 more, less obvious ones” and see what comes up.

Day 7: plan a project

I want to plan [project] with a deadline of [deadline] and a team of [who is involved].
Build a plan with the goal, stages, a week-by-week schedule, the resources needed
and the main risks. Use a table for the schedule.
At the end, list the 3 decisions I need to make before starting.

What to watch: the plan is a draft. Use it as a starting point to talk with your team, not as the final version.

After the week

Look back and answer three questions:

1. Which task did AI save you the most time on?
2. Which one did it get in the way of, or need a lot of fixing?
3. Which prompt is worth keeping?

Keep the prompts that worked. In the prompt builder you can build, generate and save your requests organized into the 4 pillars, and they stay saved in your browser for next time.