Scholarship Genie

Demo Result

Inspiration

What it does

Project Story — Scholarship Genie About the Project

Scholarship Genie is an AI-powered web application that helps students understand scholarship documents quickly and accurately. Many scholarship notifications are long, technical, and difficult to interpret. Our system uses Google Gemini AI to convert these documents into clear summaries, eligibility insights, multilingual explanations, and interactive answers. The goal is to reduce confusion and improve access to financial opportunities for students.

Inspiration

The inspiration came from a common student problem: important scholarship information exists, but it is buried inside lengthy PDFs and official circulars. Many students skip applying simply because they cannot fully understand the requirements. We wanted to build an intelligent assistant that reads these documents, explains them simply, and gives personalized guidance. The idea was to make opportunity discovery faster and more inclusive.

What We Learned

During development, we learned several practical and technical lessons:

How to connect AI APIs with a real web application workflow

How prompt design affects summary quality and reasoning accuracy

How to extract and clean text from PDF files

How to build modular backend routes for different AI tasks

How to design user flows that clearly demonstrate AI value

How to manage token limits and response structure

We also learned that good AI results depend on both clean input data and well-crafted prompts.

How We Built It

We built the project using a simple and scalable web stack:

Frontend: React interface for upload, profile input, summary view, and chat

Backend: Node.js with Express for API routing

AI Engine: Google Gemini for summarization, reasoning, translation, and Q&A

Document Processing: PDF text extraction before AI processing

Database: Cloud storage for documents and summaries

Notifications: Telegram bot for scholarship deadline reminders

Each feature — summarization, eligibility checking, translation, and Q&A — is implemented as a separate backend endpoint connected to the AI model.

Challenges We Faced

We encountered multiple practical challenges while building the system:

Cleaning noisy text extracted from PDFs

Structuring prompts to get consistent bullet-point summaries

Making eligibility reasoning explainable instead of generic

Handling large documents within model token limits

Keeping the interface simple while showing multiple features

Ensuring fast response time for demo purposes

We solved these by trimming inputs, structuring prompts carefully, and modularizing the backend logic.

Outcome

The final prototype demonstrates how AI can transform complex scholarship documents into personalized, multilingual, and interactive guidance. Students can upload a document, get a summary, check eligibility, ask questions, and receive reminders. The same architecture can be extended to government schemes, admissions, and other document-heavy services.