Teacher AI

AI-powered tool that turns teachers’ notes into custom practice exams. Automatically formats quizzes for students, saving prep time and improving learning through personalized, ready-to-use tests.

Comment

Inspiration

We wanted to build something that allows students to practice and study more efficiently and guided by their own strengths and weaknesses.

What it does

Teacher AI uses artificial intelligence to turn teachers’ notes, slides, or lesson materials into automatically formatted practice exams. The system generates multiple question types, organizes them by topic, and provides ready-to-use tests for students to take online.

How we built it

We successfully built our adaptive AI tutor by working across three synchronized tracks. Our Frontend Team quickly designed the user interface, leveraging Tailwind CSS for a fast, responsive design that included the login system and interactive forms. Meanwhile, the Backend Team laid the stable foundation using Flask and secured the environment by fixing tricky setup errors like the Pylance dependency issue. This team created a clean, two-file structure (app.py for routing and ai_service.py for logic) and integrated our data persistence via a Mock Firestore. Crucially, we connected the entire system to the Gemini AI service. Our final system is fully functional: it uses PyPDF2 to ingest context from uploaded notes and then uses that material to generate grounded, personalized lessons and structured JSON quizzes, proving we have a ready-to-demo adaptive learning MVP.

Challenges we ran into

We faced difficulties ensuring the AI-generated questions were accurate and aligned with lesson objectives.

Accomplishments that we're proud of

We successfully built a working prototype that can generate well-structured exams directly from teacher-provided notes.

What we learned

We learned how to integrate AI with educational tools and design a user-friendly interface that fits into real classroom workflow.

What's next for Teacher AI

We plan to add student analytics, grading automation, and export options to platforms like Google Classroom. In the future, we aim to include adaptive learning features that personalize quizzes based on student performance.

Share this project:

Updates