Embedded Files
Background
As the president of the Business Analytics Club, I had the unique opportunity to lead an AI Chatbot pilot project for the College of Business at Oregon State University. I was responsible for recruiting two teams of students to develop different versions of the AI Chatbot envisioned by the COB. Successfully, I recruited two student teams, managing one of them as the project manager. The goal of the AI Chatbot was to develop a full-stack web application to assist first-year COB students by answering FAQs, providing information about OSU resources, and helping with assignments.
The process
Our main goal was to build a robust and versatile LLM model, an admin file management page, and a chat page with a nice looking frontend. We divided the team into four main groups: AI, frontend, backend, and documentation. We developed our webpage to showcase two different AI models (GPT and LLaMA) using Retrieval Augmented Generation (RAG). We also implemented an interface for admins to upload and manage documents used for RAG. I was mainly involved in project management and AI development, including hosting the model, writing the code for querying GPT and LLaMA, and setting up RAG in our chatbot. Below is a short GIF showcasing our AI chatbot.
Result
I believe our team succeeded in achieving our goals: building a high-quality product, expanding our portfolio, and applying our learning in practice. I also really enjoyed the opportunity to work on an interesting project with my friends.
What I learned
Depened understanding of LLM's
Project management and task delegation
Software development management
Tools/Frameworks I used: Python (Flask, LlamaIndex) , Notion.so, Hugging Face Inference Endpoints, Github, GPT, And Meta LLaMa 3
Google Sites
Report abuse