ScholarshipHub

• 

  Dashboard

Inspiration

My friend found this scholarship. $100K. Fully funded. To study abroad.

When I asked how he found it, he was like "I just... searched for hours."

Four hours to find ONE scholarship.

Most of my friends don't even know these exist. And the ones who try? They get rejected. Not because they're not smart. But because they don't know what to write.

I got frustrated. Like actually frustrated. Why is this so hard?

So I thought: What if I just... built something?

What It Does

Okay so basically three things:

1. Finds your matches You tell it about yourself. Age, country, what you want to study, your English level. Boom. You get 5 scholarships that actually fit you. Not 500 random ones that waste your time.

2. Reviews your documents Upload your CV. AI reads it and tells you what's good and what's not. Not like "make it better lol." Like actually useful feedback. "Add numbers here." "Show impact instead of just saying you did stuff."

3. Tracks your progress You upload Draft 1. Score: 6/10. You make changes. Draft 2. Score: 8/10. You see yourself getting better. That's motivating. You keep going instead of giving up.

Also there's like a checklist for each scholarship so you don't miss anything stupid.

How We Built It

I built this solo. In under 7 days. That's the crazy part.

I used AI tools instead of hand-coding everything. That's not lazy. That's smart. Why spend 3 weeks coding when you can spend 7 days building with AI?

First thing I did: Survey 44 real students. Asked them: "What's actually hard about scholarships?" Listened to their answers. Built based on that, not based on what I THOUGHT they needed.

Then I built it:

• Used OpenCode (AI coding assistant) to generate the app structure
• I directed it: "Build these 9 screens. Use this design. Connect navigation."
• It generated code. I reviewed it. Tested it. Fixed what was broken.
• Used Claude AI for the document review feature (the smart part)
• Deployed to Vercel (literally 5 minutes)

Tested with real people. They said "This works" or "This doesn't work." I fixed it. Did it again.

No hand-coding thousands of lines. Just smart use of AI tools.

Challenges We Ran Into

The AI feedback was trash at first. Like generic. "Make it better." That's not helpful. So I spent hours rewriting the prompt. "Claude, look for THIS. Give feedback like THIS. Be specific." Eventually it worked. Took like 5 tries.

The matching algorithm almost broke me. You're ranking thousands of scholarships by fit. Age, country, education level, funding type, field of study, language. If you weight one thing wrong, everything falls apart. Tested with real people. They'd say "This doesn't match me." So I adjusted. Did it again. And again.

Time was tight. I'm a Grade 10 student. I have school. I have other stuff. But using AI tools let me move FAST.

And figuring out: How do I build this so it helps without doing the work FOR them? Like the AI shouldn't WRITE their application. It should COACH them. That distinction is actually important.

Accomplishments That We're Proud Of

Built a working app in under 7 days using AI tools. Most people talk about building things. I actually built it.

Got 44 real students to say they'd use this. 91% interested. That's not me guessing. That's real validation.

Made the AI feedback actually useful. Not generic. Specific. Actionable. That took work.

Got 234 real scholarships in the database. Real research. Real data.

Proved that you don't need to be a traditional coder to build something that works. AI tools + direction + testing = shipping fast.

What We Learned

Validate first. Talk to people BEFORE you build. That shaped everything. If I'd just coded without listening, I'd have built the wrong thing.

AI tools are powerful. But you have to direct them. Review them. Test them. You can't just ask and hope. You have to be involved.

Speed matters more than perfection. I could have spent weeks polishing. But getting a working product in 7 days and getting feedback was smarter.

The real problem isn't technology. It's access. A student in rural Egypt shouldn't have fewer opportunities than a student in Cairo just because they don't know the system. That's the actual problem technology solves.

Moving fast beats being perfect.

What's Next for ScholarshipHub

Invite those 44 people to actually USE it. See what they think. See what breaks.

Get 10 real students using it consistently. Learn from them.

Add way more scholarships. I have 234. Need thousands.

Build what they ask for, not what I think they need.

Eventually make it so they can apply directly from the app instead of jumping between tabs.

Real goal: Students actually winning scholarships with this. That's when I know it actually worked.

Built With

• css
• next.js
• opencode
• orm
• prisma
• shadcn/ui
• supabase
• tailwind
• typescript
• vercel