NextSchool

Landing page
Research case page
About page

Inspiration

When I was trying to apply for a master’s degree, I was overwhelmed. I didn’t know where to search for scholarships, constantly missed deadlines, and struggled to understand the different requirements for various programs. After a lot of late nights, disorganized notes, and frantic Google searches, I eventually figured it out — but it took way more time and effort than it should have.

That journey inspired NextSchool. I wanted to build a platform that helps aspiring postgrad students — graduates, professionals, anyone — plan and execute strong applications without the confusion and time-wasting. I wish I had something like it when I was applying.

What it does

NextSchool is an AI-powered platform that simplifies the graduate school application process by bringing everything a student needs into one intelligent space. With NextSchool, you can:

Discover and compare schools and scholarships that match your goals
Understand eligibility criteria and program requirements with ease
Stay informed with curated social posts and real-time content related to your research interests
Follow a clear, step-by-step roadmap through every stage of your application
Learn how to write a compelling Statement of Purpose (SOP)
Draft your SOP effectively using an AI-powered generator

It's designed to reduce research time and help users stay organized and confident throughout their application journey.

How I built it

I had only 5 days before the submission deadline, so I moved fast and focused on what mattered most. I chose tools that allowed rapid development while still giving me flexibility.

Frontend: Next.js
Styling: Tailwind CSS
Database: PostgreSQL
AI Integration: Perplexity's Sonar API

Challenges I ran into

One of the biggest challenges was ensuring Sonar’s AI output came in a proper JSON format. No matter how I prompted it, I kept getting inconsistencies. I experimented with different prompt structures, tried embedding the format instructions in user messages, and even looked into LangChain. But I wasn’t confident I could use it within the scope of this project, and I’m still new to that ecosystem.

Eventually, moving the JSON instructions into the system message helped stabilize the output — but it cost me valuable time and slowed development.

Accomplishments that I'm proud of

Built a fully working MVP in just a few days
Delivered real value for a problem I personally experienced
Integrated an AI-powered SOP generator
Created a clean, intuitive UI that’s easy to navigate
Brought together various resources into a single streamlined tool

What I learned

Sometimes, a problem you’ve personally faced is the best place to build from — empathy drives great product thinking.
Working under pressure teaches you to prioritize well and move quickly.
Prompting AI effectively is a skill in itself — and small changes in system/user message structure can make a big difference.
Developer experience matters — choosing familiar tools helped me ship faster and more confidently.