Demos:
Stage I Report:
CS 402 Week 4 Presentation (PDF)
Stage II Report:
Stage III Report:
Stage IV Report:
Senior Design Final Demo (PDF)
Client Meetings:
- February 13th, 2:00 pm-3:30 pm: Discussed the strengths and weaknesses of various LLMs and AI platforms.
- February 21st, 2:00 pm-3:30 pm: Make final decision on which LLM model to use and begin working on the n8n workflow.
- February 28th, 2:00 pm-3:30 pm: Parse through a metadata file provided by the library and decide on using Qdrant for our vector database.
- March 7th, 2:00 pm-3:30 pm: Switch n8n workflow template and begin utilizing Google Gemini instead of OpenAI's chat models.
- March 29th, 2:00 pm-3:30 pm: Split the project into parts, with one part using n8n as the backend and another using Python as the backend, and compare the results from both backends.
- April 4th, 2:00 pm-3:30 pm: Increase the amount of documents processed, begin utilizing Google services for storage and deployment.
- April 11th, 2:00 pm-3:30 pm: Begin testing the accuracy of the project by utilizing a "golden set" of documents with known desired answers.
- April 21st, 2:00 pm-3:30 pm: Finalize RAG testing set and begin working on Google Cloud based deployment.