In partnership with
If a recruiter asked you tomorrow to show what you can do with AI, what would you open?
A college student spent nearly two years building an AI video editor with a friend, and they stopped working on it at 13,000 users because they couldn't figure out how to grow it.
But that editor still got him two job offers at ₹50 lakh a year, and he didn't get either of them through a job portal.
The first came from a startup founder who came across it on Twitter. For the second, he picked around 30 startups building similar products and emailed them himself.
13 wrote back, and within a week a Y Combinator-backed one had made him an offer.
Notes on how candidates use AI have tripled for technical roles and quadrupled for everyone else.
Those founders could open the student's editor and judge his work for themselves, which is a lot more than an AI certificate on LinkedIn lets a hiring manager do.
Fewer seats for freshers
And interviewers are starting to check whether you understand what you've built.
Engineering leaders writing in LeadDev have suggested asking candidates to walk through a project and pick apart what the AI gave them.
So if AI built the whole thing and you can't explain how it works, that's going to come out in the interview.
A project without the coding
Rachit Sharma is a BBA student from Kaithal who says he has never written code, but he built his portfolio site with AI coding agents anyway.
He used Claude to plan each step and a coding agent called OpenCode to write the code.
He also wrote down the rules he set for them, like one small task at a time and nothing going live until the checks passed, and he describes his own job as saying no when something looks wrong.
You can get to something like that in a week if you give it an hour or two a day.
Your 7-day challenge
Day 1. Find a problem someone has.
Pick a field you want to work in and ask someone in it which repetitive task they hate doing every week, like a shop owner matching UPI payments every night or a recruiter sorting applications. Write down who has the problem and how long it takes them today, and if you can't name anyone who has it, pick a different problem.
Day 2. Decide what working looks like.
Write a short description of what your tool should do and 10 test cases it has to pass, including a few hard ones like a message in Hinglish or a missing value.
Days 3 and 4. Build it.
Use AI to build one small piece at a time, whether that's a Google Sheet with formulas or a simple web app. Save every version that works, and keep a note of what you asked the AI for and what you changed.
Day 5. Break your own project.
Run all 10 test cases and mark each pass or fail. There's usually a spot where the AI claimed something worked when it didn't, so find it, fix it, rerun the test and write down how you caught it.
Day 6. Let someone use it.
Ask the person from Day 1 to try it, and see whether a 45-minute task dropped to 10 and what confused them along the way. Then record a 90-second screen walkthrough showing what it does and one mistake you caught.
Day 7. Do what the student did.
Post the project and walkthrough on LinkedIn or X, then email the founder or hiring manager at 10 companies building something related. Open with the problem you solved and who you solved it for, because that's what they'll read first.
The Bottom Line
"Proficient in AI tools" is going the way of "Proficient in MS Office", the kind of line a recruiter skims right past.
That student's editor never found its growth, and it still got him hired.
Yours only needs one person whose week got easier and one mistake you caught along the way, and then you'll have something to open the next time someone asks what you can do with AI.
Hit reply with your answer to today's poll, and tell me what you're going to build.
Until next time,
Vaibhav 🤝🏻
If you read till here, you might find this interesting
#AD 1
Your Work May Qualify You for New Immigration Options
You’ve built systems, shipped products, and solved hard technical problems. That record may be the foundation of a U.S. Visa or Green Card case.
Manifest Law helps software engineers and technology professionals assess talent-based pathways like the O-1, EB-1, and EB-2 NIW.
Take our five-minute quiz to share your experience and goals, and an immigration professional will help you see which options may fit and what to take on next.
Attorney Advertising
#AD 2
Solana’s Price Crashed. Its Network Didn’t.
Wallets are returning, stablecoin supply is rising and institutional money is still flowing in. Our free guide explains why we’re accumulating SOL now, plus the key metrics and risks to watch.
