School AI Copilot

• 

  Landing page

Inspiration

As a student, I wanted a simpler way to study from my own notes without switching between multiple tools.

What it does

School AI Copilot lets students upload a PDF and:

• Ask questions about their notes
• Generate quizzes
• Create summaries
• Make flashcards
• Generate a 7-day study plan

How I built it

I built it using Python, Streamlit, LangChain, Chroma, Hugging Face Embeddings, and Groq. The application uses a Retrieval-Augmented Generation (RAG) pipeline to retrieve relevant sections from uploaded documents before generating a response.

Challenges I ran into

My main challenge was making sure the AI used the uploaded study material rather than unrelated information. I also had to understand how document chunking, embeddings, vector search, and LLMs work together.

Accomplishments I'm proud of

I'm proud that I independently built a working AI study assistant that combines several useful study tools into one application.

What I learned

I learned how RAG systems, embeddings, vector databases, document retrieval, and LLMs work together. I also learned how to turn a Python-based AI project into a usable web application with Streamlit.

What's next for School AI Copilot

I would like to improve response accuracy, add clearer references to source material, and make the study tools more personalized for individual students.

Built With

• chroma
• groq
• hugging-face-embeddings
• langchain
• python
• streamlit