B10 · Beginner Capstone: AI for Social Good

Project Briefing

The capstone is a final project that brings together everything in the Beginner level. Working alone or in a small team, you will identify a real problem in your community, design an AI-based solution using the AI project cycle, build a simple prototype with a no-code tool, review it for fairness and safety, and present a proposal. You are not expected to build a finished product; you are expected to think it through carefully and honestly.

8 min Beginner Jump to the activity

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

Definition

You have learned what AI is, how it learns, how to use it safely and how a project moves from problem to deployment. Now it is your turn: choose a real problem around you and design an AI idea to help.

The capstone is a final project that brings together everything in the Beginner level. Working alone or in a small team, you will identify a real problem in your community, design an AI-based solution using the AI project cycle, build a simple prototype with a no-code tool, review it for fairness and safety, and present a proposal. You are not expected to build a finished product; you are expected to think it through carefully and honestly.

In simple words

It is like planning a science fair project, but about AI and a real need. You find something that bothers people, imagine how AI could help, make a small working model, check it is fair and safe, and explain your idea to others.

Why it matters

Applying ideas to a real problem is the best way to understand them deeply. The capstone also gives you something to show, a project that demonstrates how you think, which matters more than any certificate.

Key terms

Capstone
A final project that combines what has been learned in a course.
Deliverable
Something you must produce and hand in, such as a proposal or a presentation.
Milestone
A checkpoint in a project, such as finishing the problem statement.
Prototype
A simple early version that shows how an idea could work.
Proposal
A document that explains a problem, a solution and a plan, to convince others it is worth doing.
Reflection
Thinking back on what you did, what worked, what did not and what you learned.

Explained in detail

The eight steps

StepWhat you produceLinks to
1. Project briefingA team, a timeline and a project journalThis page
2. Identify a real problemA 4Ws canvas and problem statementProblem scoping
3. Users, benefits and risksA stakeholder map and risk listDesign thinking, ethics
4. Plan the dataA system map and data planData acquisition and exploration
5. Build a prototypeA working no-code model and test resultsModelling, evaluation
6. Fairness and safety reviewA checklist with findings and fixesResponsible AI
7. Create the proposalA short written proposalData storytelling
8. Present and reflectA presentation and a reflectionCommunication

Our running example: Team Swachh

Throughout this module, you will follow Team Swachh, three students in a small town who noticed that public garbage bins near their market often overflow for days, attracting stray animals and spreading smells, while other bins are emptied when half full. Their idea: residents send a photo of a bin, an AI model classifies it as "overflowing" or "fine", and the municipal team uses a daily map of overflowing bins to plan pickups. Each page shows how the team handled that step, including their mistakes, and then asks you to do the same for your own project.

Choosing a project area

  • Environment: waste, water, trees, air quality, animals.
  • Health: reminders for medicines, healthy food in the canteen, first-aid information.
  • Education: helping younger students read, finding lost library books, learning in local languages.
  • Safety: road crossings, scam awareness for elders, reporting broken streetlights.
  • Inclusion: tools for people with disabilities, help for people who cannot read.

Working well

Keep a project journal from day one: a running record, on paper or in a shared document, of decisions, evidence, problems and changes. Divide work fairly, meet regularly and agree how decisions are made. Set milestones for each step. Above all, remember that the goal is careful, honest thinking. A modest idea, well tested and honestly described, is worth more than an impressive idea with no evidence.

You will not need to write any programs for this capstone. Everything can be done with observation, interviews, paper, a spreadsheet and a free no-code tool in a web browser. What you do need is curiosity, patience and honesty.

How it works

  1. Form a team of two to four, or decide to work alone.
  2. Start a project journal.
  3. List three or four problem areas that interest you.
  4. Agree on a timeline with a milestone for each of the eight steps.
  5. Decide who will do what, and how you will make decisions.

Try it yourself

How Team Swachh got started

Three friends walk past the market every day on their way to school.

The capstone in order

Put the capstone steps in order.

  1. 1Identify a real problem
  2. 2Study users, benefits and risks
  3. 3Review fairness and safety
  4. 4Plan the data
  5. 5Write the proposal and present it
  6. 6Build a prototype

Common mistakes

  • Choosing a problem you cannot observe or people you cannot reach.
  • Skipping the project journal and losing track of decisions.
  • Aiming for an impressive idea instead of a well-tested, honest one.

Check yourself

Question 1

1. What is the main goal of the capstone?

Question 2

2. What is a project journal for?

Question 3

3. Which project idea is the best starting point?

Summary

  • The capstone applies everything from the Beginner level to a real problem.
  • It has eight steps, following the AI project cycle with ethics built in.
  • Keep a journal, plan milestones and value honest, well-tested thinking.

Finished reading?

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