What is the problem being addressed?
Iowa State University has a digital repository that holds multiple papers such as thesis and dissertations, as well as creative components. Currently, the digital repository has filters to make the process of finding a specific paper easier. In the current digital repository, papers can be filtered by date, author, department, etc. It also has a search input that allows users to type in a specific topic and get results based on it. Although these features are helpful, it is limited. Sometimes users want more specific information about a paper, or ask detailed questions about multiple papers to find what they all have in common, or perhaps just a summary of what a paper talks about. With the current user interface and functionality, all the previous actions mentioned are not possible.
What is your project idea and how will it work (what are its components etc)
In order to solve the problem addressed above, we implemented a RAG system that provides all these features using Artificial Intelligence. Users will still have all the research papers at their disposal, and will be able to ask detailed questions about certain topics and get concise responses, depending on the quality of the prompt.
The frontend of the product is implemented through TypeScript and React to provide a modern User Interface, and uses OpenAI as the AI Agent that the user interacts with. The backend is implemented by using a webhook call to the N8N backend.