Track B
The AI Skills Track
Everything a student can legally build, ship, certify and be paid for before they turn eighteen — so that on their eighteenth birthday the business is already built and only needs switching on.
What a sixteen-year-old may legally do, and what they may not
This track is designed around a hard legal boundary rather than pretending it does not exist. Everything inside the boundary is taught fully; nothing outside it is promised at all.
Learning every AI tool that an adult freelancer uses. There is no age restriction on the software itself.
Building and publishing a portfolio — real work, live on the internet, under the student's own name.
Free industry certifications. Google and HubSpot set no minimum age. These sit on the profile for life.
Paid local work in rupees, into the student's own bank account under the State Bank's teenager account framework introduced in April 2026.
International freelance platforms. They require the account holder to be eighteen and verify identity documents. There is no consent route around it, and we will not use a parent's identity.
The 18th Birthday Switch-On — included, with no expiry
Whenever a Track B student turns eighteen — three months after the course ends or eighteen months after — they come back for a free session with the instructor. Together they open the platform account, port the finished portfolio across, complete the tax and banking registrations, and send the first proposals that same week. A Track B student does not start from zero at eighteen. They start from a finished portfolio, three certifications and a year of practice.
What this track does not claim
Track B makes no international income claim of any kind, and no member of staff may make one on its behalf. Local rupee work is real and is encouraged, but it is not guaranteed either. What is guaranteed is the portfolio, the certifications, the shipped work and the Switch-On session.
Four phases across twelve weeks
Fluency
Phase 1 — Fluency
What AI actually is
- Plain-language foundations: what a large language model is, how it is trained, what it can and cannot do, and why it sometimes states things confidently and wrongly.
- The model landscape — Claude, ChatGPT, Gemini, and open-source alternatives — and how to pick the right one for a task instead of using one for everything.
- Responsible use, honestly taught: academic integrity, disclosure, bias, and where a student's own judgement must override the machine.
Prompting as a real skill
- The six-step prompt formula — task, context, references, evaluate, iterate, format — practised until it is automatic.
- Why vague prompts produce vague work, demonstrated live on the students' own attempts.
- Building a personal prompt library, and custom assistants that behave like a specialist rather than a search box.
AI as a study and research partner
- Research that holds up: using AI to find and check sources rather than to invent them. Reading a technical document with AI assistance.
- AI for their own O and A-Levels — revision, explanation, past-paper practice, essay structure. This is the module that makes parents comfortable, because it improves the schoolwork they are already paying for.
- Presentations and documents produced to a professional standard.
Craft — the visual skills that sell
This is the heart of Track B, and it is deliberately weighted towards visual work. Visual skills are the fastest for a young person to become genuinely good at, the easiest for a client to judge without reading English, and they are the instructor's own specialism — nine years of commercial product visualisation and AI image work.
Phase 2 — Craft
Thinking like a designer
- Developing an AI mindset for design: where AI gives real leverage and where human judgement still decides. Recognising repetitive production tasks worth automating.
- Prompt engineering specifically for design: starting with a clear action verb — generate, brainstorm, draft, critique — then task, context and references.
- The fundamentals that AI cannot supply for you: hierarchy, contrast, spacing, and type.
Images — the instructor's own craft
- Commercial-grade image generation: Midjourney, Nano Banana and Stable Diffusion. Prompt structure, aspect and composition control, style consistency across a set.
- Art direction rather than slot-machine pulling: choosing a reference, controlling the output, and iterating deliberately toward a brief.
- Product visualisation — placing a real local product into a scene that looks photographed. This is billable work anywhere in the world.
Video, voice and short-form content
- AI video: Veo, Kling and Runway. What is genuinely usable commercially today and what still looks like a toy.
- Voice and avatars — ElevenLabs and HeyGen — plus sound and editing in CapCut.
- Short-form content that holds attention: thumbnails, hooks, captions, and the eight-step production workflow.
Words that sell, and the basics of being found
- Copy that persuades: headlines, offers, and the difference between describing a product and describing what it does for the buyer.
- Social and ad content produced at speed, on brand, without sounding machine-written.
- Search fundamentals — what SEO actually is, why businesses pay for it, and enough of the vocabulary to sit the HubSpot certification in Week 11.
Ship — real work, for a real business
From Week 8 the class stops making exercises and starts making deliverables. Each student takes on one real local business — a family shop, an uncle's practice, a business on Khayaban-e-Sehar, or the academy itself — and delivers properly, to a deadline.
Phase 3 — Ship
Build a website without writing code
- From prompt to first-draft layout, then concept, system, remix, upgrade — the same build sequence taught in Track A.
- Why AI-generated design usually looks cheap, and the specific moves that fix it: depth, texture, real content, restraint.
- Publishing it live on a real domain that the student owns.
Your first automation
- How systems talk to each other: triggers, logic, actions — and just enough JSON to stop it being frightening.
- Building one genuinely useful automation end to end and letting it run without supervision.
- Where automation is worth money to a business, using the four pillars of a process worth automating.
Deliver it, and be paid for it
- Handling a real client: taking a brief, agreeing a scope, setting a deadline and saying no politely to scope creep.
- Quoting local work honestly at a student rate, and invoicing it.
- Delivery, revisions, and asking for a written testimonial — which becomes the first line of social proof on the portfolio.
Ready — the business, built and waiting
Phase 4 — Ready
Certifications and your own bank account
- Three free certifications, completed in class: Google Ads Search, Google Analytics and HubSpot SEO. None has a minimum age. All three are free. All three sit on a professional profile permanently.
- The credibility paragraph they unlock, written for each student in their own words.
- Opening a bank account in the student's own name under the State Bank's teenager account framework — the student's account, their debit card, their app, with a guardian indemnity as the regulator requires. Local earnings go here.
The portfolio, and the profile that waits
- The portfolio site finished properly: five pieces, each written as a case study — the problem, what you did, what changed.
- The platform profile drafted in full and held. Headline, overview, rate, portfolio pieces and certifications are all written and saved — but nothing is published and no account is opened, because the student is not yet eighteen. It waits for them.
- A dated personal plan: what to practise between now and the eighteenth birthday, and the standing invitation to the Switch-On session.
What your child actually walks out with
And one thing that is not on any syllabus: a sixteen-year-old who has delivered real work to a real business, been paid for it, and been asked to revise it, is a different person in a university interview than one who has not.