Problem
Students often need a specific answer that is buried across a university website. Manually navigating pages can be slow and makes relevant information harder to find.
Implementation
I built an LLM chatbot using Python, LangChain, and Reflex. Retrieval augmented generation grounds its answers in university documents rather than relying only on the model’s general knowledge.
Technical decisions
- Retrieved university content to provide context for generated answers.
- Used LangChain to organize the retrieval and generation workflow.
- Built the application interface with Reflex.
- Added basic guardrails and response validation to keep answers within the intended content scope.
Results
Students can ask questions through a chat interface that retrieves information from university documents.