Computer science is, by almost every measure, the most competitive undergraduate major in the world. At top U.S. universities, CS acceptance rates are often half the overall rate. Stanford's School of Engineering admits around 5% of applicants. Carnegie Mellon's School of Computer Science is even more selective. MIT, Princeton, and Caltech are in the same range.

So what actually works?

Borderless is a platform of over 560 stories from international students who studied abroad. We went through all of them and identified 78 who were admitted to study computer science or a closely related major (software engineering, data science, computer engineering) at universities ranging from MIT and Princeton to Grinnell and Berea. These students came from over 25 countries, including Uzbekistan, Kazakhstan, Egypt, Armenia, Kyrgyzstan, Bangladesh, India, Syria, Belarus, Thailand, Brazil, Spain, Japan, Cambodia, Greece, Rwanda, Poland, Ukraine, and beyond.

This isn't advice from admissions consultants. This is what accepted CS applicants really did, really wrote about, and really built. Here are the seven things that worked.

1. Build something real that people actually use

This was the single clearest signal across every strong application. The students who got into top CS programs didn't list online courses or certificates. They built functioning products that solved actual problems, and other people used them.

The form varied widely. Some built apps: browser extensions, Telegram bots, farm management tools with real sensors. Some launched businesses: web design agencies, ride-hailing platforms for women drivers, freelance development practices. Some created tools for specific communities: an AI assistant in an underserved language, a program that automated document translation at a journalism agency, an app that summarized privacy policies for people who never read them.

What made these projects compelling wasn't technical sophistication. It was that they were real. They had users, or clients, or revenue. Saidafzal from Uzbekistan, who got a full ride to Colby College, completed over 300 freelance projects starting at age 12. Dilafruza, also from Uzbekistan, co-founded a ride-hailing startup generating $2,000 per month in revenue and got a $380,000 full ride to Rice University.

The takeaway is clear: build something. Don't wait for a class assignment or a hackathon prompt. Find a problem that frustrates you or someone you know, and write the code that fixes it. The bar isn't "impressive." The bar is "real."

2. Combine CS with something unexpected

Almost nobody who got into a top program positioned themselves as "just a CS person." The students who stood out had a second area of deep interest, and the intersection of that interest with CS became their spike.

These combinations were often surprising. CS and linguistics. CS and disability rights. CS and animal welfare. CS and cultural preservation. CS and factory manufacturing. The second area was rarely another STEM field. It was personal, rooted in the student's identity, community, or lived experience.

Aikhan from Kyrgyzstan built his entire Princeton application around using NLP to create a translator that captures cultural nuance in the Kyrgyz language. Alba from Spain used AI and neural networks to study Galician, a minority language in her region, and got a full ride to Yale.

This makes strategic sense. When tens of thousands of applicants say they love computer science, the differentiator isn't how much you love it. It's what you do with it that nobody else does. The intersection creates a story that is impossible to replicate, because it comes from your specific background, your specific community, and your specific obsessions. An admissions officer reading 500 CS applications will remember the student who used machine learning to preserve an endangered language. They won't remember the twentieth student who built a to-do app.

3. If you're self-taught, frame it as a strength

You might assume that students admitted to top CS programs had access to AP Computer Science, well-funded robotics teams, or international school STEM curricula. Most didn't. The majority of the CS students in our dataset learned to code on their own, outside of any formal curriculum, through YouTube, Coursera, freelance work, or sheer trial and error.

This is partly a reflection of where these students came from. Public schools in Samarkand, Tokmok, Aleppo, and rural Armenia don't offer AP CS. But what's interesting is how the strongest applicants turned this disadvantage into a narrative asset. Rather than apologizing for the lack of formal training, they made the gap between their resources and their output the central argument of their application.

Victoria from Belarus, who got a full ride to Wesleyan, explicitly framed it this way: she showed admissions that she accomplished projects at the level of students from well-resourced international schools (Chrome extensions, published research, chatbot development) despite having zero programming classes at her regular school. The contrast was the story.

For applicants without access to formal CS education, the advice is counterintuitive: don't try to hide it. Lean into it. Show what you built with nothing, and let the admissions committee do the math on what you'll build with the resources of a top university behind you.

4. Write a personal essay, not a technical one

Across 78 CS applicants, almost nobody wrote their main essay about code, algorithms, or technology. The essays that worked were deeply personal, about identity, family, culture, or a formative experience that had nothing to do with programming.

One student wrote about secretly rescuing three abandoned kittens and hiding them in his basement. Another used a steering wheel metaphor, opening with himself as a child on his father's lap nearly crashing the car and ending with him driving confidently years later. Another wrote a prose fiction piece about falling in love with three girls, which turned out to be about learning three languages. One repurposed a love poem about Jupiter and Saturn into a UChicago supplemental. Another connected algorithms and string theory to a story about playing the violin.

The pattern is unmistakable: CS skills belong in your activities list and supplemental essays. Your main essay is about who you are as a person. The strongest essays revealed something about the student's character, worldview, or identity that couldn't be communicated through a resume or a GitHub profile.

As one Colby student put it: he applied as a CS major, but his essay about rescuing cats read like political science, and he was fine with that. The lesson: don't force your major into your essay. Let the essay be about you, and let the rest of your application be about CS.

5. Teach CS to others

This was one of the most underrated signals in the data. A disproportionate number of accepted CS students had taught coding to other people, and it wasn't a minor line item on their activities list. For many, it was one of their most important extracurriculars.

The formats varied: founding Python teaching clubs, running web development workshops through organizations like Girls in Tech, teaching senior citizens to use smartphones, volunteering as coding instructors during gap years, answering hundreds of questions on Stack Overflow. Some taught formally through nonprofits and educational platforms. Others taught informally through online communities and peer mentoring.

Aikhan from Kyrgyzstan, who got into Princeton, founded a tech education platform with 8,000+ social media followers and 100+ volunteers, and separately taught over 100 senior citizens how to use apps like WhatsApp, inspired by watching his grandmother struggle with technology.

Teaching CS is strategically powerful because it accomplishes two things at once. It demonstrates genuine mastery (you can't teach neural networks to high schoolers if you don't deeply understand them yourself) and it signals community impact, which is what admissions committees at need-blind schools weight most heavily. If you're looking for a single extracurricular that strengthens a CS application more than almost anything else, teaching coding to others is it.

6. Do research at your local university

Research experience appeared across the dataset, but not from the places you'd expect. The students who published papers or conducted meaningful CS research did it at local universities in their home countries, through remote research programs, or through community organizations, not at MIT or Stanford labs.

One student published a paper on treating PTSD using VR glasses at his local medical university in Samarkand. He got the opportunity by walking into a professor's office hours and asking. Another published two AI research papers while in high school in Kyrgyzstan. Another worked with an NYU professor remotely through Pioneer Academics. Several used EducationUSA's Competitive College Club to produce scientific publications.

Admissions committees care about the output (a published paper, a conference presentation, a demonstrable result) not the prestige of where you did it. A research paper from Samarkand State Medical University carries the same weight on your application as one from a famous lab, as long as the work is real and the output is tangible.

For students in countries without obvious research infrastructure, the playbook is straightforward: approach a professor at your nearest university, propose a project that combines CS with a problem you care about, and commit to producing a publishable result. Remote programs like Pioneer Academics and Lumiere Research also provide structured paths to published research with faculty mentorship.

7. Prioritize projects over competitions

Competition results appeared in many profiles, but they were rarely the centerpiece of an application. Only a handful of students in our dataset had international-level medals, and many of the strongest applicants had no major competition results at all.

The students who got into Princeton, Duke, Wesleyan, and Colby with full rides had zero competition medals between them. What they had instead were functioning products, published research, teaching portfolios, and freelance experience. Meanwhile, the one student with the strongest competition record (IMO silver and EGMO gold medals from Kazakhstan) got into MIT but didn't even apply as a CS major. She planned to study math and discovered CS later.

This doesn't mean competitions are useless. National-level results in informatics or robotics olympiads clearly help. But projects consistently outperform medals as the anchor of a CS application. The reason is simple: competitions show you can solve predefined problems under time pressure. Projects show you can identify real problems, design solutions, build them, and ship them to actual users. The second set of skills is closer to what CS programs actually teach, and what admissions committees are looking for.

If you have strong competition results, include them. But if you don't, don't panic. Spend that time building something instead.

The bottom line

Across 78 CS students from 25+ countries, admitted to schools ranging from MIT and Stanford to Colby and Berea, these are the seven things that consistently worked:

1. Build real products that other people actually use.
2. Combine CS with a non-obvious second interest and let that intersection become your defining story.
3. Treat self-teaching as an asset, not a gap to apologize for.
4. Write deeply personal essays that reveal your character, not your technical skills.
5. Teach CS to others to demonstrate mastery and community impact at the same time.
6. Do research at accessible institutions and produce tangible outputs.
7. Prioritize building over competing by investing time in projects that ship.

Every student in this dataset connected CS to something bigger than CS itself. It was never about the code. It was about what the code made possible: translating endangered languages, protecting people from misinformation, teaching grandmothers to use smartphones, making factories safer, helping women drivers earn a living.

That's the real answer to how you stand out in the most competitive major. You don't just show them you can code. You show them why you code.

Start your own journey

If you're an ambitious international student, your background is not a limitation. It's your story. And universities with full scholarships are actively looking for students like you.

Borderless is a free, AI-powered platform that helps international students navigate admissions, find scholarships, and build competitive applications. No expensive consultants. No gatekeeping. Just the tools and guidance you need.

Read more stories from students who did it, and start writing yours.