Prepmind AI

• 

  AI Practice Page

• 

  Landing page

• 

  Ai Tutor page

Inspiration

I was inspired to build PrepMind after seeing friends spend thousands on books and expensive SAT tutoring centers like Kabir Prep. Many felt pressured to join, even if they couldn’t afford it. I wanted to change that so anyone could practice anywhere and generate unlimited questions, and prepare without cost or endless searching online for good resources.

What it does

PrepMind Ai(prepmind.org) is a by students for students free web app that helps democratize SAT, ACT, AP exam prep. Pick exam and subject, generate infinite multiple choice questions, with step by step explanations. Use a voice/text Ai tutor for instant help. An open source, privacy respecting, community first, sustainable platform. Built to close gaps in access to quality exam prep.

How we built it

The stack is practical and lightweight. The UI is built with React and Vite, styled with Tailwind CSS, and tested to work well in dark and light modes. I code in VS Code, push changes to GitHub, deploy the site on Render, and manage the prepmind.org domain and HTTPS through IONOS. A small Node and Express service formats requests, validates inputs, and returns consistent results. No user accounts are required and no profile data is stored at rest.

Challenges we ran into

One major challenge was optimizing how questions were generated. At first, each request caused long delays and timeouts. I solved this by creating a batching system that grouped question prompts and processed them efficiently through the server. It reduced monthly costs from $103.58 to $0.58 and cut generation time by 56%, keeping it free and accessible for everyone.

Accomplishments that we're proud of

The Challenge also pushed me to communicate results, not just intentions. I can point to a cost drop from 103.58 dollars to 0.58 dollars and a 56 percent improvement in average generation time, with a focused path improving from 5.0 seconds per question to about 1.6 seconds. Sharing those outcomes helped me see engineering as a process with clear problems, concrete fixes, and measurable gains. Most of all, I learned that access matters. When a tool is fast, free, and respectful of privacy, more students stick with it and keep practicing. That is the impact I want, and it will guide how I continue to grow PrepMind.

What we learned

Participating in this hackathon taught me that real impact comes from solving problems that matter. I learned how to design scalable code, manage APIs efficiently, and balance technical goals with community needs. It showed me that accessibility, not complexity, is what makes technology meaningful.

What's next for Prepmind AI

For Prepmind AI 2.0, I plan to add offline access for areas with limited internet, multilingual support with natural voices, and faster response handling for smoother performance. I also want to broaden its reach beyond test prep, adapting the same accessible learning for classrooms. Bringing the student centered, high quality learning experience to all levels of education.

Built With

• api
• css
• github
• html
• ionos
• javascript
• typescript
• vscode