Track A
The Earning Track
From no experience to a verified international profile with live proposals going out under supervision — and a first paid client targeted inside the twelve weeks.
Four phases, and only one of them is about tools
Most AI courses stop at the tools and leave the student holding a skill nobody has paid them for. This one spends a third of its hours on the part that actually converts skill into money: finding the client, writing the proposal, and closing the call.
The rule that governs Phase 4
No student spends a single rupee of platform credit on a proposal the instructor has not read. Proposal credits are released weekly, against reviewed work, with a hard weekly cap. This is simultaneously the strongest cost control in the programme and the core of what the family is paying for.
Foundations and the money rail
The rail is built in the first fortnight, long before there is any money to put through it, so that when a first contract closes in Week 11 nothing is improvised.
Phase 1 — The Rail
What AI actually is, and where the money is
- Session 1 — The landscape without the hype: what a large language model is and is not, open versus closed models, tokens and context windows, and why "the worst AI you will ever use is today's". Essential terminology, in plain language.
- Session 2 — The five business models AI actually pays for, and where a beginner in Karachi realistically enters. Choosing a lane: AI design and websites, or AI automation. Nobody leaves this session undecided.
- Session 3 — Reading the market. Live walkthrough of real job postings on the instructor's own account: what clients ask for, what they pay, and the difference between a job post and the problem behind it.
Prompting properly, and building the money rail
- Session 4 — The six-step prompt formula: task, context, references, evaluate, iterate, format. Why most people's prompts fail. System prompt versus user prompt. Building a reusable prompt library you will use for twelve weeks.
- Session 5 — Custom assistants and projects. Making a model behave like a specialist rather than a search engine. Reactive and dynamic prompting.
- Session 6 — The money rail. FBR registration and filer status; PSEB freelancer registration; the bank appointment; foreign-currency retention; Payoneer; and the tax position in writing.
The tax position students are taught, in one line
Registered with PSEB and filing, a freelancer pays 0.25% of export proceeds as a final tax, deducted automatically by the bank when the money arrives. On eight hundred dollars a month that is two dollars. Unregistered and not filing, the same student pays 2% — eight times as much. This single session pays for a meaningful part of the fee.
One sellable skill, built to a real deliverable
Students commit to one of two lanes in Week 1 and go deep. Both lanes converge on the same thing by Week 6: a finished piece of work good enough that a stranger would pay for it.
Phase 2 — The Skill
Lane 1: from prompt to a designed page · Lane 2: data, JSON and APIs
- Lane 1 — Zero to a first AI-built site: prompt to first-draft layout, then concept, system, remix and upgrade. Visual hierarchy, colour and type for interfaces people scan. Why "AI slop" looks like AI slop, and the specific moves that lift it to award-level: depth, texture, real content.
- Lane 2 — How data actually moves: JSON structure, APIs and HTTP requests, authentication and credentials, reading documentation without fear. This is the week that separates people who can automate from people who can only follow tutorials.
Lane 1: the full build · Lane 2: your first working automation
- Lane 1 — A complete multi-page site for a real local business — chosen in Week 3 and briefed properly. Copy written with AI and edited by a human. Images generated and art-directed. Deployed live on a real domain.
- Lane 2 — First end-to-end automation: trigger, logic, action. Nodes versus tools, credentials, pinning test data, conditional logic and error handling. Building something that runs without you watching it.
Going beyond the obvious build
- Lane 1 — Motion, 3D and the details that make a page feel expensive. Cloning and reverse-engineering a site you admire as a learning method. Speed: how to deliver in days what agencies quote in weeks.
- Lane 2 — AI agents proper: the agent node, chat models, choosing a model for a job, memory, text classifiers and information extractors, structured output parsing.
- Both lanes — Human-in-the-loop and quality control. What you put in front of a client and what you never do.
Process mapping: seeing the work a client cannot describe
- Session — Intro to process mapping. The four pillars of a process worth automating. How to walk into any business and find three things worth money within twenty minutes.
- Session — Return on investment, stated in the client's language. Turning "this saves nine hours a week" into a number and a price.
- Session — Productising: turning a bespoke build into a repeatable offer with a fixed price — the difference between being hired and being booked.
Proof, and a profile written to be found
Phase 3 — The Proof
Portfolio, case studies and the video that closes
- Session — A portfolio is not a gallery. Writing each piece as a case study: the client's problem, what you did, what changed. Three pieces, properly written.
- Session — Your own portfolio site, live on your own domain, with your name on it.
- Session — The Loom proposal. Recording a two-minute video that shows the client you already understand their problem — the single highest-converting technique taught in this course, and the one that lets a Pakistani freelancer beat a native English speaker on the same job.
The profile build — line by line, together
- Session — The headline is a search title, not a job title. Photograph, overview paragraph, and the credibility sentence. Rate setting: why you start below the category average and how the ladder to $50 works.
- Session — Certifications. Three, all free, all completed in class: Google Ads Search, Google Analytics and HubSpot SEO. Then the credibility paragraph they unlock.
- Session — Identity verification, the rail tested end to end with a live test transaction, and profile review by the instructor before anything goes live.
Seven ways in that almost nobody uses
The proposal strategies below are the instructor's own, developed across 102 contracts. They are the reason a new freelancer from Karachi can win against a crowded field, and they are taught nowhere else.
Bid on the problem, not the job
Every job post is a symptom, not a diagnosis. Name the real problem in the first two lines and the client stops reading the other twenty proposals.
The portfolio approach
Do a small piece of the work before you are hired and attach it. Costs an hour; removes all doubt.
The jobs-in-progress interceptor
Find work already being done badly and offer to finish it properly. The least competitive listing on the platform is the one someone has already messed up.
First one free
Deliberate, scoped, and time-boxed — never open-ended. Converts scepticism into a paid second job.
The platform-to-LinkedIn bridge
Turning a one-off contract into a direct, long-term relationship the legitimate way, once the platform relationship allows it.
Reverse hunting
Stop waiting for postings. Identify the businesses that need the work, then create the opening yourself.
Automation stack-up
Turning one delivered automation into three more for the same client. This is how a $300 job becomes a $997-a-month retainer.
Written by hand, every time
Proposals are written by the student, not generated. Clients can tell, and the ones who cannot are not the clients you want.
Supervised hunting
Four weeks in which the student stops learning and starts working, with the instructor reading everything before it goes out. This is the part of the programme that cannot be recorded, cannot be scaled, and is the reason the batch is capped at twelve.
Phase 4 — The Hunt
First proposals out
- Job selection: which listings to ignore, and why most beginners burn their credits on jobs they were never going to win.
- Writing four proposals in class, each reviewed line by line before submission.
- Quota: four proposals this week. Deliberately low. Fewer and better beats more and worse, and the quota falls as quality rises.
The call
- Handling the first client message. Response speed as a competitive weapon — one of the instructor's own reviews reads "within five minutes we were on a Google Meet."
- Live mock discovery calls, recorded and critiqued. Scoping the job on the call. What to say when you do not know the answer.
- Quoting: fixed price versus hourly, milestones, and the three-day risk-free close.
Deliver, and get the review that compounds
- Delivering a first contract properly: communication cadence, scope discipline, and handling the client who asks for more.
- Why the first review is worth more than the first payment, and exactly how to earn a five-star one.
- Contingency: if no contract has landed by this week, the instructor subcontracts a real, paid piece of live client work to the strongest students — genuine work, genuine money, a genuine review.
From one job to an income
- The retainer conversation: converting a completed project into a monthly arrangement. The automation stack-up in practice.
- The rate ladder: when to raise your price, by how much, and the exact words to use with an existing client.
- Ninety-day plan written by each student, plus the alumni group and the standing monthly review call.
The honest version
Every outcome listed in the "You leave with" lines above is guaranteed, because every one of them is within the instructor's control. A first paid client is the target of the twelve weeks and it is what the method is built to produce — but it depends on the student's own hours and effort, and it is not guaranteed.
What is guaranteed in writing: attend ninety percent of sessions and submit every assigned proposal, and if you finish without a verified profile, a live portfolio, PSEB registration and a working payment account, the fee is refunded in full.