StudyPilot – AI Assistant Chatbot for Students

Inspiration

Students often have lecture slides and notes but struggle to turn them into an effective study plan. We wanted to build an AI assistant that could take existing study material and transform it into an interactive learning experience instead of simply giving students another chatbot.

What it does

StudyPilot allows students to upload their lecture slides or notes as a PDF. Then ask AI to:

Extracts and identifies key topics from the material
Generates 10 MCQs with explanations
Provides an interactive quiz
Calculates the student's score
Identifies weak areas based on quiz performance
Generates a personalized revision plan
Provides additional practice questions
Allows students to ask questions about their study material through an AI chat

How we built it

We built StudyPilot using Python and Streamlit for the application interface.

For the AI capabilities, we integrated the Google Gemini API to analyze study material, generate questions, evaluate performance, create revision plans, and answer student questions.

We used PyMuPDF to extract text from uploaded PDF lecture material and Python logic to process MCQs, calculate quiz scores, and maintain the student's session data.

Tech Stack

Python
Streamlit
Google Gemini API
PyMuPDF
python-dotenv

Challenges we ran into

One of the main challenges was reliably converting AI-generated MCQs into structured data that the application could use. The AI response needed to follow a specific JSON structure so the quiz could be generated correctly.

We also had to design the application flow so that uploaded material, quiz answers, scores, and study plans remained connected throughout the student's session.

Accomplishments that we're proud of

We built a complete AI-powered study workflow rather than just a question-answer chatbot.

A student can go from:

Lecture PDF → AI Analysis → Quiz → Score → Weak Areas → Personalized Study Plan → AI Chat

This makes the system more focused on actual exam preparation and personalized learning.

What we learned

We learned how to integrate a generative AI model into a real application and connect AI-generated outputs with application logic.

We also gained experience with PDF text extraction, structured AI responses, quiz evaluation, session-state management in Streamlit, and designing an AI workflow around a real student problem.

What's next

We want to expand StudyPilot into a more complete AI assistant for students.

Future improvements could include:

Support for multiple PDFs and subjects
Long-term student progress tracking
Adaptive quizzes based on previous performance
Automatic flashcard generation
Exam countdown and study scheduling
Voice-based interaction
More advanced retrieval so the AI can answer questions directly from large collections of study material
A personalized AI tutor that continuously adapts to each student's learning needs

Vision

Our goal is to make StudyPilot more than an AI chatbot — we want it to become a personal AI study assistant that understands a student's material, measures their progress, and helps them prepare more effectively.