B10 · Beginner Capstone: AI for Social Good

Creating the Proposal

A project proposal is a short, clear document that explains a problem, the evidence for it, a proposed solution, how it was tested, its benefits, risks and limits, and what is being asked for next. A good proposal is honest, evidence-based and written for its reader.

15 min Beginner Jump to the activity

What you need: Nothing. This page starts from the beginning, with no programming and no formulas.

Definition

The municipal commissioner has five minutes to read about Team Swachh's idea. What should be on the first page so that she understands the problem, believes the evidence and knows what to decide?

A project proposal is a short, clear document that explains a problem, the evidence for it, a proposed solution, how it was tested, its benefits, risks and limits, and what is being asked for next. A good proposal is honest, evidence-based and written for its reader.

In simple words

A proposal is like asking your parents for a new bicycle, but done well. You explain why you need it, show you have thought about cost and safety, admit what could go wrong, and say exactly what you are asking for.

Why it matters

Good ideas need to convince people with the power to act on them. Writing a proposal also forces you to organise your thinking and to separate what you know from what you hope.

Key terms

Proposal
A document that argues for a solution and asks for a decision or support.
Executive summary
A short opening section giving the whole proposal in a few sentences.
Audience
The person or group who will read the proposal and make the decision.
Limitation
Something the solution cannot do, or has not been tested on.
Next steps
What should happen after the proposal, such as a pilot.
Ask
The specific decision or support you are requesting.

Explained in detail

The structure

SectionWhat it saysTeam Swachh
Executive summaryThe whole proposal in four sentencesProblem, solution, key result, the ask
The problemWho, what, where, why, with evidenceTwo of six bins overflowed on 9 of 14 days
The solutionHow it works, in plain words, with a diagramPhoto → classifier → daily map for the supervisor
Evidence it worksPrototype test results, including failures90 per cent accurate; misses cut from 4 to 1 in 30
BenefitsWho gains and howCleaner streets, fewer wasted truck trips
Risks and safeguardsWhat could go wrong and how it is handledFace filter, minimum-visit rule, phone reporting
LimitationsWhat is untested or cannot be doneRainy season untested; old bins 86 versus 90 per cent
The ask and next stepsThe specific decision requestedA one-month pilot in one ward, with the supervisor's involvement

Write for your reader

The commissioner is busy and not an AI expert. So the executive summary comes first and stands on its own. Technical words are explained or avoided: "the system sorts photos into full and not full" rather than "binary image classification". One clear chart, such as overflow days per bin before and after, says more than a page of numbers. Every claim is backed by evidence, and the source of the evidence is stated.

Team Swachh's executive summary

"Two of the six public bins near our market overflowed on most days we checked, while others stayed half empty, because the sanitation team cannot see which bins are full. We built a prototype that sorts residents' photos of bins into 'overflowing' and 'fine' and was 90 per cent accurate on photos it had not seen, with safeguards for privacy and fairness. It has not yet been tested in the rainy season. We ask for a one-month pilot in Ward 4, planned with the sanitation supervisor."

Honesty makes it stronger

It can be tempting to hide limitations. But a reader who finds a hidden problem stops trusting everything else. A reader who sees limitations stated openly, with a plan to address them, trusts the rest more. If you used AI tools to help draft or polish the proposal, follow your school's rules and say so. The thinking, evidence and decisions must be your own.

How it works

  1. Identify your reader and what decision they can make.
  2. Write each section using your journal: problem, solution, evidence, benefits, risks, limitations, the ask.
  3. Add one clear chart and one simple diagram.
  4. Write the executive summary last, and put it first.
  5. Ask someone outside the team to read it and say what they understood.

Try it yourself

The first draft that was too long

Team Swachh writes their first proposal draft.

Order the proposal

Put the sections in a sensible order.

  1. 1The problem and evidence
  2. 2Executive summary
  3. 3The ask and next steps
  4. 4Risks, safeguards and limitations
  5. 5The solution
  6. 6Test results

Common mistakes

  • Burying the main result deep inside the document.
  • Using technical words the reader does not know.
  • Hiding limitations, or ending without a specific ask.

Check yourself

Question 1

1. Where should the executive summary go?

Question 2

2. Why include limitations?

Question 3

3. Which is a good "ask"?

Summary

  • A proposal explains the problem, solution, evidence, benefits, risks, limitations and ask.
  • Write for the reader: summary first, plain words, one clear chart.
  • State limitations honestly and end with a specific ask.

Finished reading?

Sources: CBSE Artificial Intelligence curriculum resources · United Nations Sustainable Development Goals