If you are on F-1 and looking for your first US role, you are in an awkward position: the thing that gets you hired is experience, and the thing you are not allowed to do much of is work. CPT needs to be tied to your programme. Internships are seasonal. Off-campus work is restricted.
Hackathons sit outside that. You can enter one this weekend without a work permit, without your DSO's sign-off, and without a company deciding you are worth a visa risk. That is a genuinely unusual gap in the rules, and it is worth more than the prize.
But most hackathon projects get written up in a way that wastes them. Here is what the software reading your resume does with them.
What a screening system sees
Before a person reads your resume, an applicant-tracking system parses it into fields. The analyzer here runs the same kinds of checks. Three of them decide whether a project entry helps you:
Does the line start with an action verb? Systems and recruiters both scan the first word of each bullet. Built, designed, deployed, automated, migrated, reduced all read as work you did. Anything else reads as context.
Is there a number in it? A bullet with no measurable outcome is the single most common weakness we flag. Not because numbers are magic, but because they are the difference between a claim and a result.
Does it avoid the phrases that drain it? Responsible for, worked on, helped with, assisted with, participated in, involved in, exposure to. Every one of those puts distance between you and the work. Participated in a 36-hour hackathon says you attended. It is the most common way a good project becomes a line nobody reads.
The same project, written two ways
Weak, and extremely common:
Participated in HackTX 2026. Worked on a team of 4 to build a web app for campus food waste. Responsible for the backend.
Nothing here is false, and nothing here is useful. Three of the drain phrases. No verb you own. No outcome.
Stronger:
Campus food-waste tracker — HackTX 2026, 36 hours, team of 4 Built the Flask backend and PostgreSQL schema handling 2,000 seeded donation records with sub-200ms query times. Integrated the Google Maps API for pickup routing. Placed 3rd of 47 teams.
Same weekend, same project. The difference is that the second version names what you built, what it handled, and how it did against a field you can measure.
Where the numbers come from when you have none
The objection is always the same: my project had no users, so I have nothing to count. You have more than you think.
- Scale of what it handled — records, rows, API calls, images, files, requests per second
- Speed — response times, build times, how long the thing you automated used to take
- The field — 3rd of 47 teams, top 10 of 200, selected from N applicants
- The constraint — 36 hours, team of 4, one weekend, no prior experience with the framework
- Scope — endpoints, models, screens, integrations
"Built in 36 hours" is a real constraint and a real number. Use it.
What you should never do is invent one. A fabricated metric is the thing most likely to end an interview badly, because the follow-up question is always "how did you measure that?"
Which hackathons are worth your weekend
Not all of them are, and the ones that pay off are not always the biggest:
- Ones where you ship something you can demo later. A live URL or a public repo outlasts the certificate.
- Ones with company sponsors in your field. Sponsor engineers judge, and they are often the same people who screen resumes. That is a warmer introduction than an application form.
- Ones where you build the thing you cannot build at work. If you want to move from data analysis into engineering, a hackathon is where you get an engineering artefact without asking anyone's permission.
Prize money is the least valuable thing on offer.
Where it goes on the page
Under a Projects section, above Education if you have no US work experience, below Experience if you do. Two or three entries, not seven. A hackathon project is not weaker than an internship on a resume — it is only weaker when it is written like an attendance record.
When you have rewritten it, run the resume through the analyzer and see whether the bullets clear the checks above. It will tell you which ones still have no verb and no number.