Quickprep.ai

Web App – AI in Education

COMPANY

Team Project

MY ROLE

Team Lead

TOOLS

Flask
Groq API
HTML CSS JS
LLama 3.3 70B

TIMELINE

2025

DESCRIPTION

Quickprep.ai is an AI-powered web application designed to help teachers generate quizzes and track student performance effortlessly. By analyzing uploaded module notes, the platform uses large language models to generate multiple-choice questions (MCQs), which students can answer online. Teachers can then monitor individual and class-wide performance through a dynamic dashboard.

CONTEXT

Creating high-quality MCQs and tracking student understanding can be time-consuming for teachers. With Quickprep.ai, we aimed to remove these bottlenecks by using AI to automate quiz generation while enabling a streamlined interface for students to test themselves and track their progress. This made the learning loop faster, easier, and more transparent for both students and teachers.

View on GitHub
Quickprep.ai interface

Challenge

How might we enable teachers to generate quizzes from existing teaching materials, while also giving them detailed insights into individual student progress—all with minimal manual work?

1.
Generate well-structured, valid MCQs from raw module notes using LLMs.
2.
Ensure the output adheres to strict JSON schema, despite LLM unpredictability.
3.
Allow students to view quizzes, attempt them, and track performance by difficulty level.

Process

The project started with conversations with educators about the pain of creating new assessments. I explored how LLMs could transform this process. We experimented with several local and open-source models before settling on a cloud-hosted LLaMA 3.3 model via Groq for its performance and ease of integration. Based on early feedback from both students and teachers, we continuously refined the UI and quiz validation logic.

1.
Gathered feedback from teachers and students on quiz challenges.
2.
Built PDF-to-MCQ conversion flow with JSON schema enforcement.
3.
Implemented two user roles with clean, functional dashboards.
4.
Iterated interface and difficulty tagging system based on classroom usage.

Solution

Quickprep.ai offers two distinct interfaces: one for teachers and one for students.

1.
Teachers can upload module notes, generate MCQs via AI, review or edit them, and track individual student performance through a detailed dashboard.
2.
Students can sign up, take quizzes, and receive real-time feedback. Each question is tagged with a difficulty label (easy, medium, hard), and students can reattempt quizzes with average score tracking.

The AI-generated questions follow a strict JSON schema, validated through a custom script. Teachers are not required to craft questions from scratch—just review and tweak.

Dashboard UI

Takeaway & Reflection

This project deepened my understanding of LLM integrations and their practical constraints in educational tools. It also reinforced how thoughtfully applied AI can relieve educators from repetitive tasks, giving them more time to focus on pedagogy.

From a product perspective, it taught me how even simple UIs—when tuned to user feedback—can powerfully serve both ends of a system: teachers and students.