Main page
Scholarship results page

Scholar Scout - Project Story

About the Project

Scholar Scout is an AI-powered scholarship matcher that helps students find scholarships tailored to their unique academic profile, background, and goals. Whether you're a top-performing undergrad in Computer Science or a first-gen student looking for financial support, Scholar Scout streamlines the search process using the intelligence of GPT.

Inspiration

We were inspired by the frustration many students feel when searching for scholarships — the process is often tedious, outdated, and filled with irrelevant results. We wanted to create something that felt personal, smart, and effortless. With the power of AI, we saw an opportunity to revolutionize how students access educational funding.

What We Learned

How to integrate OpenAI's GPT API into a real-time web app
Techniques for parsing and rendering dynamic JSON responses beautifully on the frontend
How to handle user input and securely send that data to an AI model for personalized output
The importance of CORS, async handling, and clean UI/UX in a responsive web application

How We Built It

Frontend: Vanilla HTML/CSS + JavaScript for simplicity and speed. Markdown files are rendered using the marked.js library.
Backend: Flask (Python) server handles POST requests, sends data to the GPT API, and returns personalized scholarship lists as JSON.
AI Engine: OpenAI's GPT-4 model formats scholarship recommendations based on GPA, ethnicity, major, and academic level.
Display: Parsed results are shown as elegant, scrollable cards with key details and application links.

Challenges We Faced

Parsing GPT Responses: Ensuring the returned data was clean and valid JSON took some trial and error.
API Rate Limits: We had to manage how often requests were sent to avoid hitting GPT rate limits during testing.
Data Accuracy: Ensuring scholarships were relevant and up-to-date required crafting precise prompts and filtering responses smartly.