CampusQuery AI

Inspiration 💡

CampusQuery AI was inspired by our own frustration as high school students trying to find simple information on school websites. Important details about sports tryouts, graduation requirements, and course prerequisites were buried across long pages and PDFs, while counselors were overwhelmed answering the same questions repeatedly. We realized the problem wasn’t missing information—it was how hard it was to find.

What it does 🤖

CampusQuery AI turns school websites into intelligent, AI-powered assistants. Parents, students, and staff can ask natural-language questions and instantly receive accurate, source-cited answers grounded in real school content, available 24/7.

How we built it 🛠️

We built CampusQuery AI using a full Retrieval-Augmented Generation (RAG) pipeline. School websites are crawled with Firecrawl, broken into chunks, embedded using OpenAI’s text-embedding-3-small model, and stored in PostgreSQL with pgvector. When a user asks a question, relevant content is retrieved via semantic search and passed to GPT-4.1-mini to generate accurate, contextual responses. The frontend is built with React, TypeScript, and Tailwind CSS, while the backend runs on serverless Edge Functions.

Challenges we ran into 🚧

We faced challenges building a real-time chatbot interface, accurately crawling dynamic school websites, scaling full-site ingestion without timeouts, and tuning semantic search thresholds for reliable answers. Learning how embeddings, similarity scores, and RAG systems work together required extensive experimentation and iteration.

Accomplishments that we're proud of 🏆

We built a production-grade AI system that schools actually want to use. We successfully implemented semantic search, accurate source citations, real-time analytics, and scalable crawling. We also validated our idea by interviewing students, parents, counselors, school foundation leaders, and district IT staff.

What we learned 📖

We learned how to think like systems engineers, balancing performance, accuracy, and usability. Beyond technical skills, we learned the importance of user-centered design, collaboration, and building technology that solves real-world problems.

What's next for CampusQuery AI 🚀

Next, we plan to add PDF document ingestion, multi-language support, proactive notifications, personalization for students, and more advanced analytics. Our goal is to evolve CampusQuery AI from a chatbot into a fully intelligent digital assistant for school communities starting with our school district.