Inspiration FAU has 90+ policy documents buried across their website. Students miss deadlines and break rules simply because the information is hard to find. We wanted to fix that.
What it does CampusBuddy is an AI chatbot that instantly answers FAU student questions about policies, academic deadlines, and campus life — always citing the exact source document.
How we built it We scraped all 90 FAU policy documents, chunked them into a ChromaDB vector database, and built a RAG pipeline using LangChain and HuggingFace embeddings. When a student asks a question, the most relevant policy chunks are retrieved and sent to Claude AI, which generates a sourced answer. The frontend is built with Streamlit.
Challenges we ran into FAU's policy website loads content via JavaScript, so standard scraping failed. We had to reverse-engineer their internal AJAX API to extract clean policy text.
Accomplishments that we're proud of Scraping and ingesting all 90 FAU policies in one day, building a full RAG pipeline from scratch, and creating a chatbot that never hallucinates — it only answers from real documents.
What we learned How to build a production-style RAG pipeline, handle dynamic web scraping, and design AI systems that cite sources instead of guessing.
What's next for Campus Buddy Connect to live FAU systems for real-time deadlines, expand to other universities, add multilingual support for international students, and build a mobile app.
Built With
- huggingface
- langchain
- python
Log in or sign up for Devpost to join the conversation.