Senior school · portfolio track
Build · Level B
Full curriculum document
Senior-school students building a portfolio before campus
Contents · 6 units
- 1Computer science from first principlesBLD-U1 · Months 1–2 · 60 GLH · Portfolio + exam
- 2The web, end to endBLD-U2 · Months 3–4 · 70 GLH · Internal portfolio
- 3Working like an engineer: Git, testing & reviewBLD-U3 · Months 5–6 · 60 GLH · Internal portfolio
- 4How machine learning worksBLD-U4 · Months 7–8 · 70 GLH · Portfolio + exam
- 5Building with AI: LLMs & responsible useBLD-U5 · Months 9–10 · 70 GLH · Internal portfolio
- 6Capstone: ship & defendBLD-U6 · Months 11–12 · 80 GLH · Portfolio + exam
Overview & approach
A program is not magic — it is data, logic and structure a human designed. Build starts by making that structure visible: decompose a problem, choose the right data shape, then write code you can explain line by line and defend in review.
The first-principles learning loop
- 1Problem
- 2Understanding
- 3Reasoning
- 4Design
- 5Implementation
- 6Testing
- 7Reflection
- 8Improvement
Every unit runs this loop, then goes round again — deeper each time.
Competencies, values & contemporary issues
7 CBE core competencies
- Communication & collaboration
- Critical thinking & problem solving
- Creativity & imagination
- Citizenship
- Digital literacy
- Learning to learn
- Self-efficacy
Values
- Integrity
- Responsibility
- Respect
- Excellence
- Unity
- Patriotism
Pertinent & contemporary issues
- Digital citizenship & online safety
- Data protection & privacy
- AI ethics, bias & fairness
- Environmental sustainability (compute cost)
- Financial literacy & entrepreneurship
- Inclusion & accessibility
- Mental well-being & healthy screen habits
- Career & life skills
The 6 units in full
BLD-U1 · Level 2 · Months 1–2 · 60 GLH · Portfolio + exam
Computer science from first principles
Strand: Foundations of software & problem-solving
Key inquiry: “When a problem looks too big to code, how do we break it down until each piece is obvious?”
You ship: A command-line tool that solves a self-chosen problem, with tests and a written design rationale.
LO1 · Data structures & algorithms
- AC 1.1 Choose an appropriate data structure and justify the choice
- AC 1.2 Analyse a solution’s efficiency in plain terms (time/space)
LO2 · Decomposition & OOP
- AC 2.1 Decompose a problem into functions/classes with clear responsibilities
- AC 2.2 Refactor working code to be simpler and explain each change
LO3 · Version control
- AC 3.1 Use Git to branch, commit and collaborate with a partner
Core competencies
- Critical thinking & problem solving
- Digital literacy
- Self-efficacy
PCIs addressed
- Data protection & privacy
- Career & life skills
Hands-on practice
- Solve a real problem two ways and compare which is clearer and faster.
- Pair-program a tool and review each other’s commits.
- Whiteboard the design before writing code; reflect on what changed.
Unit grading (Pass · Merit · Distinction)
| Pass | Solves guided problems and uses Git. |
|---|---|
| Merit | Designs and refactors independently with sound choices. |
| Distinction | Delivers reusable, tested, well-documented tooling. |
BLD-U2 · Level 3 · Months 3–4 · 70 GLH · Internal portfolio
The web, end to end
Strand: Full-stack engineering
Before this unit: Comfortable decomposing problems and using Git (Unit 1).
Key inquiry: “What actually happens between a click in a browser and a change in a database?”
You ship: A deployed full-stack app (frontend + API + database) with a live URL.
LO1 · Frontend
- AC 1.1 Build an accessible, responsive interface with modern JS/React
- AC 1.2 Manage state and handle user input and errors
LO2 · Backend & data
- AC 2.1 Design a REST API and a simple database schema
- AC 2.2 Connect frontend to backend and deploy the result
Core competencies
- Critical thinking & problem solving
- Communication & collaboration
- Creativity & imagination
PCIs addressed
- Data protection & privacy
- Inclusion & accessibility
Hands-on practice
- Trace one request from click to database and back, and diagram it.
- Ship a small app to a live URL and share it for feedback.
- Add accessibility (labels, contrast, keyboard) and test it.
Unit grading (Pass · Merit · Distinction)
| Pass | Builds and deploys a basic full-stack app. |
|---|---|
| Merit | Adds robust state, errors and accessibility independently. |
| Distinction | Ships a polished, accessible product real users can use. |
BLD-U3 · Level 3 · Months 5–6 · 60 GLH · Internal portfolio
Working like an engineer: Git, testing & review
Strand: Technology engineering practice
Before this unit: A deployed full-stack app (Unit 2).
Key inquiry: “How do professional teams keep a growing codebase correct when many people change it at once?”
You ship: A tested feature merged through pull requests, with code review and CI passing.
LO1 · Testing & CI
- AC 1.1 Write unit tests and run them automatically on every push
- AC 1.2 Read a failing test and fix the code behind it
LO2 · Collaboration
- AC 2.1 Branch, open a pull request and review a peer’s code
- AC 2.2 Resolve merge conflicts and keep a clean history
Core competencies
- Communication & collaboration
- Critical thinking & problem solving
- Self-efficacy
PCIs addressed
- Career & life skills
- Data protection & privacy
Hands-on practice
- Add tests to an untested project until you trust a change.
- Review a classmate’s pull request and leave constructive comments.
- Set up CI so tests run automatically on every push.
Unit grading (Pass · Merit · Distinction)
| Pass | Writes tests and uses pull requests with support. |
|---|---|
| Merit | Reviews code and keeps CI green independently. |
| Distinction | Runs a professional review-and-test workflow others can follow. |
BLD-U4 · Level 3 · Months 7–8 · 70 GLH · Portfolio + exam
How machine learning works
Strand: Introduction to AI & ML
Before this unit: Confident in Python and building apps (Units 1–3).
Key inquiry: “A model isn’t programmed with rules — so how does it “learn”, and how do we know it learned the right thing?”
You ship: A trained classifier with an honest evaluation and a short write-up of its limits.
LO1 · Data & models
- AC 1.1 Prepare data and train a baseline model with Python ML libraries
- AC 1.2 Explain supervised learning and what the model is optimising
LO2 · Evaluation & ethics
- AC 2.1 Read accuracy/precision honestly and spot over/under-fitting
- AC 2.2 Check the model for bias and report the finding
Core competencies
- Critical thinking & problem solving
- Learning to learn
- Citizenship
PCIs addressed
- AI ethics, bias & fairness
- Environmental sustainability (compute cost)
Hands-on practice
- Train a classifier, then try to break your own evaluation.
- Audit predictions across groups and present what you found.
- Estimate the energy cost of training and discuss trade-offs.
Unit grading (Pass · Merit · Distinction)
| Pass | Trains and evaluates a model against a given metric. |
|---|---|
| Merit | Selects metrics and diagnoses failure independently. |
| Distinction | Delivers a fairness-checked model with a defensible story. |
BLD-U5 · Level 3 · Months 9–10 · 70 GLH · Internal portfolio
Building with AI: LLMs & responsible use
Strand: Applied AI for builders
Before this unit: A trained model and a deployed app (Units 2–4).
Key inquiry: “How do we add a language-model feature that is genuinely useful — and safe when it is wrong?”
You ship: An app with an LLM feature, guardrails, and an honest evaluation of where it fails.
LO1 · Building with LLMs
- AC 1.1 Add an LLM feature with clear prompts and structured output
- AC 1.2 Ground answers in real data and handle wrong answers gracefully
LO2 · Responsible use
- AC 2.1 Test for harmful or biased output and add guardrails
- AC 2.2 Explain the feature’s limits honestly to a user
Core competencies
- Creativity & imagination
- Critical thinking & problem solving
- Citizenship
PCIs addressed
- AI ethics, bias & fairness
- Data protection & privacy
Hands-on practice
- Add an LLM feature to an existing app and make its output reliable.
- Red-team your own feature for unsafe or wrong answers.
- Write a short “how this AI can be wrong” note for users.
Unit grading (Pass · Merit · Distinction)
| Pass | Adds a working LLM feature with support. |
|---|---|
| Merit | Grounds output and adds guardrails independently. |
| Distinction | Ships a genuinely useful, safety-checked AI feature. |
BLD-U6 · Level 3 · Months 11–12 · 80 GLH · Portfolio + exam
Capstone: ship & defend
Strand: Synoptic project
Before this unit: Completed the software, web, ML and AI units (Units 1–5).
Key inquiry: “What separates a class exercise from a product someone would actually choose to use?”
You ship: A complete, deployed project (software, web or AI) with docs, and a portfolio-ready presentation.
LO1 · Build & document
- AC 1.1 Scope, build and deploy a complete solution to a real problem
- AC 1.2 Document it so a stranger could run and understand it
LO2 · Present
- AC 2.1 Demonstrate the project and explain every design decision
- AC 2.2 Reflect on what you would improve and why
Core competencies
- Self-efficacy
- Communication & collaboration
- Creativity & imagination
PCIs addressed
- Career & life skills
- Financial literacy & entrepreneurship
Hands-on practice
- Choose a problem you care about and scope it ruthlessly.
- Deploy publicly and gather feedback from real users.
- Present to a mentor panel; defend your choices.
Unit grading (Pass · Merit · Distinction)
| Pass | Ships a working, documented project. |
|---|---|
| Merit | Validates with users and iterates independently. |
| Distinction | Delivers a portfolio-grade product with a compelling defence. |
What you walk away with
A portfolio of 6 real, shipped artefacts — one per unit, assessed to standard.
- 1 A command-line tool that solves a self-chosen problem, with tests and a written design rationale.
- 2 A deployed full-stack app (frontend + API + database) with a live URL.
- 3 A tested feature merged through pull requests, with code review and CI passing.
- 4 A trained classifier with an honest evaluation and a short write-up of its limits.
- 5 An app with an LLM feature, guardrails, and an honest evaluation of where it fails.
- 6 A complete, deployed project (software, web or AI) with docs, and a portfolio-ready presentation.
Assessment rubric (CBE & Cambridge)
| CBE level | Cambridge grade | What the evidence shows |
|---|---|---|
| Below expectation | Not yet Pass | Cannot yet demonstrate the outcome unaided; work is incomplete or does not run. Needs re-teaching and another attempt. |
| Approaching expectation | Pass | Meets every assessment criterion with guidance. The artefact works for the happy path and the learner can explain what it does. |
| Meeting expectation | Merit | Meets all criteria independently, handles edge cases, evaluates the result, and justifies design decisions from first principles. |
| Exceeding expectation | Distinction | Ships a production-grade artefact to real users, critiques trade-offs (cost, latency, safety), and extends beyond the brief with original work. |