CBSE Class 12 · Code 843 · Last verified 3 November 2026
CBSE Class 12 AI capstone project (843): a 2026-27 guide
In short
The Class 12 capstone in Artificial Intelligence (843) is a team project that tackles a real problem linked to a Sustainable Development Goal, and it carries 25 of the 50 practical marks: 15 for the project, 6 for its documentation and 4 for a video that must last exactly three minutes. The viva draws on it too. This guide turns CBSE’s 2026-27 rules and its project guidelines into a plan a team of three to five can follow.
- Teams have at least 3 and at most 5 students, and the project must be aligned with an SDG.
- Marks: capstone 15, documentation 6, video 4, out of a 50-mark practical half; the 5-mark viva covers the capstone and the practical file.
- The video must be exactly 3 minutes and cover six set points.
- Teams may build in Python or with no-code and low-code tools such as Orange Data Mining.
What are CBSE’s capstone rules for 2026-27?
- Work in a group of three to five students.
- Align the project with one of the UN Sustainable Development Goals.
- Complete the project and its documentation in Class 12, following the “Project Guidelines” at the start of CBSE’s AI Projects Cookbook.
- Make a video that covers the problem statement, the SDG it addresses, the AI concepts, domains or algorithms used, how the project works, a conclusion, and an acknowledgement to the teacher.
- Build with Python or with a no-code or low-code platform such as Orange Data Mining; CBSE leaves the choice to the students.
From: Artificial Intelligence (Sub. Code 843), Class XII: Curriculum (Job role: AI Assistant), Capstone Project guidelines, p. 9
“Video of the Capstone Project should be exactly of 3 minutes duration.”
How are the Class 12 practical marks split?
| Component | Marks |
|---|---|
| Capstone project | 15 |
| Project documentation | 6 |
| Project video | 4 |
| Practical file | 10 |
| Lab test in Python and Orange Data Mining | 10 |
| Viva voce on the capstone project and the practical file | 5 |
| Part C total | 50 |
From: Artificial Intelligence (Sub. Code 843), Class XII: Curriculum (Job role: AI Assistant), scheme of units, p. 2
Theory Unit 2, Data Science Methodology, is subtitled “An Analytic Approach to Capstone Project” and carries 8 theory marks. One of its learning outcomes is to build the methodology’s steps into the capstone, so the project doubles as revision for the theory paper.
From: Artificial Intelligence (Sub. Code 843), Class XII: Curriculum (Job role: AI Assistant), Unit 2, p. 4
Which stages does a capstone go through?
| Stage | Suggested hours | Data science methodology link |
|---|---|---|
| Prepare for the project | 1 | — |
| Form a team | 1 | — |
| Define the problem (SDG-aligned) | 4 | Business understanding |
| Understand your users | 4 | Business understanding |
| Brainstorm the solution | 3 | Analytic approach |
| Design your solution | 1 | Analytic approach, data requirements |
| Identify and collect data | 3 | Data collection, understanding and preparation |
| Build your prototype | 4 | Modelling |
| Test your solution | 2 | Evaluation, feedback |
| Reflect as a team | 0.5 | Feedback |
| Reflect individually | 0.5 | — |
| Create a video pitch | 6 | Deployment (presenting the solution) |
| Total | 30 |
In Class 11 the documentation covers the first six stages; in Class 12 the team completes the project and its documentation.
From: Artificial Intelligence (Sub. Code 843), Class XI: Curriculum (Job role: AI Assistant), Project Guidelines chart, p. 8; AI Projects Cookbook for Classes XI & XII (Subject Code 843), Project Guidelines
CBSE’s cookbook also suggests team roles, so that the work is shared: a project leader who plans and keeps time, a data expert, a researcher, a designer, a prototype builder or coder, a tester, a communications lead who assembles the documentation, and a video producer. In a smaller team, people take on more than one role. Record who held which role in the documentation.
From: AI Projects Cookbook for Classes XI & XII (Subject Code 843), Project Guidelines, “Form a team”
How do you choose a good capstone problem?
Start close to home
Look for a problem in your school or neighbourhood that you can observe and test with real users, not a global problem you can only read about.
Check that AI genuinely helps
The cookbook’s test is whether the solution needs something AI is good at: recognising images, understanding speech or text, finding patterns in a lot of data, or reading sensors. If ordinary rules would do the job, pick another problem.
Write a measurable problem statement
Name the user, what they need to do and how you will know it worked, for example: “help our canteen staff sort waste correctly so that less food waste goes into the dry bin, measured over two weeks.”
Make sure you can get the data ethically
Prefer data you create yourself, such as photos of objects, or open public datasets. Do not collect classmates’ faces, names or personal details, and get permission before using any records.
Rank the ideas
Place each idea on a grid of value to the user against ease of building, and pick one that is valuable and buildable in the time you have.
From: AI Projects Cookbook for Classes XI & XII (Subject Code 843), Project Guidelines, stages 1–5
Which project ideas fit the rules?
| Idea | SDG | AI technique | Data | Tool |
|---|---|---|---|---|
| Leaf-disease spotter for the school garden | 2 Zero Hunger | Image classification | Your own photos of healthy and diseased leaves | Teachable Machine or Orange image analytics |
| Canteen waste sorter | 12 Responsible Consumption | Image classification | Your own photos of paper, plastic and food waste | Orange or Teachable Machine |
| Water-use forecaster for a hostel or school | 6 Clean Water | Regression | Monthly meter readings, with the school’s permission | Python (pandas, scikit-learn) |
| Library book recommender | 4 Quality Education | Association rules | Borrowing records with names removed, with permission | Orange |
| Event feedback mood board | 4 Quality Education | Sentiment analysis | Anonymous feedback-form comments | Orange text mining |
| Air-quality explainer | 3 Good Health | Classification | Open public air-quality data for your city | Python or Orange |
Whatever you build, report its accuracy honestly, with the metric that fits: precision and recall for classifiers, RMSE for regression.
What should the documentation include?
- Title page: project name, team members and each person’s roles.
- Problem statement and the SDG it serves, with why it matters to your users.
- Users: who they are, what you observed or asked them, and an empathy map.
- Ideas considered and why you chose this one (your priority grid).
- Data: where it came from, permissions, features, how much, and how you cleaned it.
- Model: the approach (for example classification), the tool, how you trained it and how you evaluated it, with the numbers.
- Testing: what users tried, what they said and what you changed.
- Ethics and limits: possible bias, privacy safeguards and where the model fails.
- Reflections: one as a team and a short one from each member.
- References for every dataset, tool and source.
From: AI Projects Cookbook for Classes XI & XII (Subject Code 843), Project Guidelines
How should the three-minute video be planned?
| Time | Part | What to show |
|---|---|---|
| 0:00–0:30 | Problem statement | The user and the problem, filmed where it happens if possible |
| 0:30–0:45 | SDG | Which goal, and the link in one sentence |
| 0:45–1:30 | AI used | Domain, algorithm and tool, with one simple diagram |
| 1:30–2:30 | How it works | A live demo on a new example, with its result |
| 2:30–2:50 | Conclusion | What worked, the accuracy you measured and the next step |
| 2:50–3:00 | Acknowledgement | Thanks to the teacher who guided you |
From: Artificial Intelligence (Sub. Code 843), Class XII: Curriculum (Job role: AI Assistant), video components, p. 9
Script it, rehearse it, time it, then trim it. A video that runs over or under breaks the “exactly three minutes” rule.
How can a team prepare for the viva?
- Every member should be able to explain the problem, the data and the model, not just their own part.
- Be ready to say why you chose your evaluation metric and what your numbers mean.
- Know the project’s limits: where it fails, what bias it could have and how you protected privacy.
- Link it to theory: name the data science methodology stage each part of the project belongs to.
From: Artificial Intelligence (Sub. Code 843), Class XII: Curriculum (Job role: AI Assistant), scheme of units, p. 2; Unit 2, p. 4
Frequently asked questions
- How many students can be in a Class 12 AI capstone group?
- At least three and at most five, according to CBSE’s 2026-27 curriculum for 843.
- How long should the capstone video be?
- Exactly three minutes. It must cover the problem statement, the SDG, the AI concepts or algorithms used, how the project works, a conclusion and an acknowledgement to the teacher.
- Can the capstone be built in Orange instead of Python?
- Yes. CBSE’s note for educators says students may choose Python or no-code and low-code platforms such as Orange Data Mining.
- How many marks is the Class 12 AI capstone worth?
- 25 of the 50 practical marks: 15 for the project, 6 for the documentation and 4 for the video. The 5-mark viva also covers the capstone.
Sources and versions
We link to the documents as their publishers host them. We read each file in full on the date shown; open the publisher's PDF to check anything on this page.
Artificial Intelligence (Sub. Code 843), Class XII: Curriculum (Job role: AI Assistant)
CBSE Department of Skill Education · Curriculum for session 2026-2027 · 12 pages · last verified 3 November 2026
Units, hours and marks are unchanged from the 2025-26 curriculum.
Artificial Intelligence (Sub. Code 843), Class XI: Curriculum (Job role: AI Assistant)
CBSE Department of Skill Education · Curriculum for session 2026-2027 · 12 pages · last verified 3 November 2026
Units, hours and marks are unchanged from the 2025-26 curriculum.
AI Projects Cookbook for Classes XI & XII (Subject Code 843)
CBSE Department of Skill Education · Edition listed for 843 on CBSE’s skill education portal, checked for 2026-27 · 43 pages · last verified 3 November 2026
The 843 curricula point to its “Project Guidelines” for the capstone process and documentation.
Related guides
Parents: start your child free
LittleAIMaster is an AI course app for Grades 6–12 with short, interactive lessons on how AI works. It is free to start, and you decide whether to upgrade.
Teachers: free AI worksheets for your class
Printable AI worksheets for Grades 6–12 that you can use in a lesson today.
LittleAIMaster is an independent learning app. It is not affiliated with or endorsed by CBSE, NCERT or any government education board. We link to the documents CBSE and NCERT publish rather than copying them, and quote them only briefly, with the source named.