How to build a coding portfolio that holds up under questioning.
A portfolio is not a gallery of screenshots. It is a set of claims about what you can do, and every one of them can be tested by a single follow-up question. Here is how to build one where that goes well.
Classes that produce portfolio-grade work
Each one ends in something built, documented and defensible rather than a certificate.
Python and AI classes for kids and teens
The usual starting point. Language fundamentals to the level where a first real project becomes possible.
View course
Vibe coding, used properly
How to build with AI assistance and still be able to explain every line, which is what an interview will test.
View course
AI, ML, Python and Java track
The longer route for students aiming at research-grade work rather than a single app.
View course
Data science course
Collecting, cleaning and interrogating data, which is most of the work in any real project.
View courseIn short
A coding portfolio is a small set of finished projects, each with a version-controlled repository, a working demonstration and a clear written explanation. MIT asks code portfolio submissions for the codebase, ideally version-controlled, plus a working demonstration by video, live link or executable. Two or three genuinely finished projects beat ten unfinished ones.
Modern Age Coders works with students to take a project all the way to that standard and then rehearses them on defending it. To talk about where your child is now, use the form below or message us on WhatsApp at +91 91233 66161.
The short version
- Two or three finished projects, not ten started ones. Reviewers open the first two and stop.
- MIT's stated requirement for code is the codebase plus a working demonstration. Build to that even if you never apply there.
- Version control history is part of the evidence. A repository created in one commit says something you cannot unsay.
- In the UK there is no portfolio slot at all for computer science, so the portfolio's job is to feed the personal statement and survive the interview.
- Waterloo requires an Admission Information Form for all its Computer Science programs and evaluates on grades and the form together, so in Canada this material is genuinely scored.
Where a portfolio is actually read
The slot varies enormously by country. Build to the strictest standard and it works everywhere.
United States: an optional formal slot
MIT accepts a Maker Portfolio through SlideRoom. Code projects must include the codebase, ideally version-controlled, and a working demonstration. MIT states portfolios are truly optional and should only be submitted when the work is significant and relevant.
Source: MIT Admissions, Creative portfoliosCanada: a required form that counts
The Admission Information Form is required for admission to all programs in Waterloo's Cheriton School of Computer Science, and the Faculty of Mathematics evaluates applicants on both grades and the form, considering courses, non-academic activities and math contests.
Source: Waterloo Cheriton School of Computer ScienceUnited Kingdom: it lives in the conversation
Oxford states extra-curricular activities do not form part of the selection criteria in any subject, while super-curricular work including programming experience helps build the picture, and advises using something you have actually done to show commitment.
Source: Oxford Department of Computer ScienceUCAS: the written version of your portfolio
From 2026 entry the personal statement is three questions within a 4,000 character total and a 350 character minimum per section, and the third question is specifically about preparation outside education.
Source: Guide to the 2026 UCAS personal statementThe eight steps
In order. Steps one to four are the work, five to eight are what makes the work legible to somebody else.
Steps 1 to 4: the work itself
Most of the time goes here1. Choose two or three projects, and cut the rest
Pick the ones you can explain best, not the ones that took longest. Everything else comes out of the portfolio entirely.
A weak third project drags down the two strong ones. Reviewers average, they do not take the maximum.
2. Take each one to genuinely finished
Finished means someone else can run it, it handles being used wrongly, and it does not have a dead button in the main flow.
Unfinished work invites the one question you cannot answer: why did you stop?
3. Use version control from the first commit
Real history, meaningful messages, work spread over the period you claim to have spent on it.
MIT asks for version-controlled code specifically. The history is evidence the work happened over time.
4. Get one real user per project
Somebody unrelated to you who used it and told you something. Fix at least one thing because of what they said.
Changing your design because of feedback is the clearest signal of engineering maturity available.
Steps 5 to 8: making it legible
Two weekends, high return5. Record a demonstration for each project
One to three minutes. Show it working, show what happens when the input is wrong, and narrate what you built rather than what it does.
A working demonstration is an explicit requirement for MIT code submissions and it is what everyone else wants anyway.
6. Write a README a stranger can follow
What it is, how to run it, what is not implemented. Assume the reader has three minutes and no context.
The README is what a reviewer reads before deciding whether to look at your code at all.
7. Write the build log
A short document per project: what you tried, what failed, what you would do differently. Written as you go, not reconstructed afterwards.
This is the document that makes the whole portfolio credible, because nobody manufactures dead ends.
8. Rehearse the defence out loud
Have someone technical question you for thirty minutes per project. Every stumble is a task on next week's list.
In the UK this is the only form the portfolio takes. Elsewhere it is still where a strong application is confirmed.
The six checks before you call it done
Run every project through these. Anything failing two or more is not ready to be in the portfolio.
Can a stranger run it?
Hand the README to someone who has never seen it and watch them try. Do not help. Fix what stopped them.
Is there a working demonstration?
Video, live link or executable. If the only artefact is a screenshot, the project is not demonstrable.
Does the history look like the story?
If you claim four months of work, the commit log should show four months of work.
Can you explain every line?
Including anything an AI assistant wrote. If you cannot explain it, either learn it or remove it.
Does it fail gracefully?
Empty input, no network, wrong file. Handling failure is the difference between a demo and software.
Is there one honest limitation stated?
Every project has one. Writing it down makes every other claim more believable.
A realistic build schedule
For a student working a few hours a week alongside school, starting from being able to code a little.
Project one, small and complete
Deliberately modest. The goal is to experience finishing something, which most students never have.
Project two, with a real user
Bigger, and put in front of someone. Iterate on what they struggled with.
Project three, the ambitious one
Only now, once finishing is a habit rather than a hope.
Documentation and demos
READMEs, build logs and recorded demonstrations for all three. Two focused weekends.
Defence rehearsal
Thirty minutes of hostile questioning per project. Fix what breaks.
Freeze and submit
Stop adding. A frozen, coherent portfolio beats one still being edited the night before.
A portfolio is a record of time spent
That is precisely why it is hard to fake, and why when a student starts changes what the finished portfolio can be.
Starting at 6 or 7, building for years
Years of commits, abandoned experiments and rebuilt projects produce a portfolio with a story. A student who started at seven has that story without ever having planned for it.
- Ages 6 to 9Block coding and logic. Finishing small things, which is the habit everything later depends on.
- Ages 10 to 13Real code in Python. First projects that another person can actually run.
- Ages 14 to 16AI and machine learning, own datasets, work that goes in front of real users.
- Ages 17 to 18Portfolio, competitions and research-grade work, with years of history behind it.
By the time it matters, the repository is years deep. That cannot be assembled in a hurry, and it is obvious to anyone who looks.
A deadline in weeks, not years
From a standing start, two or three genuinely finished projects plus documentation and rehearsal takes about ten to eleven months. Compressing it further mostly removes the parts that made it credible.
- Weeks 1 to 2Scope honestly against the time left. Most ideas get cut down here, and that is the useful part.
- Weeks 3 to 6Build the smallest version that genuinely works, with the repository open from day one.
- Weeks 7 to 8One real user, then fix what tripped them up.
- Final weekREADME, recorded demo, and rehearsal for being questioned about it.
Smaller and finished beats ambitious and broken. We will tell you before you start if the deadline does not allow honest work.
Send us what already exists. The most useful first conversation is usually an honest read on what is worth finishing and what to drop.
What we do with students on this
Almost every student we meet has started more projects than they have finished. The work is in the finishing.
Two or three projects, actually finished
A mentor takes the student through the whole arc: choosing what is worth keeping, taking it to a standard a stranger can run, writing the documentation as they go, and then sitting on the other side of the table asking the questions an interviewer will ask.
- An honest cut of what belongs in the portfolio and what does not
- Each surviving project taken to runnable by a stranger
- READMEs and build logs written during the work, not reconstructed after
- Demonstrations recorded to the one to three minute standard
- Thirty minutes of hostile questioning per project, then a fix list
If there is not much built yet, these come first.
Python and AI classes for kids and teens
Beginner
Vibe coding, used properly
All levels
AI, ML, Python and Java track
Advanced
What portfolio work costs
Portfolio work is taught in normal classes at normal fees. There is no separate portfolio package and no per-project charge.
Group class
Batches of 5–8. Good for building the fundamentals.
₹1,499 per monthMini batch
3–4 students. More airtime per student.
₹2,999 per monthOne to one
1:1. The usual choice for a project with a deadline.
₹7,500 per monthPrices are per month and shown per region. Hosting a demo and a public repository is free on the services students normally use. See full pricing for every subject and tier.
Portfolios our students have built
Real projects from Modern Age Coders students, published on Student Labs.
Pracy, a coding practice and quiz platform
Multi-user, with question banks and scoring. Big enough that architecture decisions had to be made and defended.
FreshTrack, a food expiry tracker
Small, finished and genuinely used. The kind of project that survives a follow-up question because it solves a problem the student actually had.
HugSelle, an AI smile generator
A narrow, well-scoped model task taken to a finished, usable interface.
SkyCast, a weather forecast app
Third-party API, error states, and what to show when the network fails. Unglamorous, and exactly what reviewers probe.
Tell us what your child has built so far
Tell us what exists and what state it is in. We reply within one business day with which projects are worth finishing and which to drop.
Want the work looked at first?
Send us a repository link. An honest read on what is worth keeping is usually the most useful first conversation.
Your details are used only to reply to your enquiry and are never shared or sold. We help students finish their own work, and we do not write code for a portfolio.
Frequently asked
What should be in a coding portfolio?
Two or three finished projects, each with a version-controlled repository, a working demonstration of one to three minutes, a README a stranger can follow, and a short build log of decisions and dead ends. MIT specifically asks code submissions for the codebase plus a working demonstration.
How many projects should a portfolio have?
Two or three. Reviewers rarely get past the first two, and a weak extra project lowers the average rather than adding to it.
Does a portfolio matter for UK university applications?
There is no portfolio slot for computer science. Oxford states extra-curricular activities are not part of the selection criteria, though subject-related work helps build the picture. The portfolio's job is to give you something concrete for the personal statement and to survive the interview.
Does it matter in Canada?
At Waterloo, yes. The Admission Information Form is required for all Cheriton School of Computer Science programs, and the Faculty of Mathematics evaluates on grades and the form together, including non-academic activities and math contests.
Should projects be on GitHub?
Yes. Version control is part of what is being assessed, and MIT asks for the codebase to be version-controlled where possible. The commit history is evidence in its own right.
What if my child used AI to write some of the code?
That is common and not disqualifying. The requirement is that they can explain every line they submit. Anything they cannot explain should be learned properly or removed, because a single follow-up question will find it.
Is a personal website necessary?
Helpful, not essential. A clean repository with a good README does most of the work. A site is a convenience for the reader rather than evidence in itself.
How long does building a portfolio take?
About ten to eleven months at a few hours a week, starting from basic coding ability. Three projects, then a month of documentation and a month of rehearsal.
Do you help students build one?
Yes, that is what our Portfolio Program does: live mentoring through the build, documentation as you go, and mock technical questioning at the end. We do not write the code and we do not offer admissions consulting.