Locally deploying an AI Chatbot using Retrieval Augmented Generation on Consumer Hardware with Custom Data from Alstom
Year: 2025
Type: dissertation
Abstract: As the use and interest of AI increases, there is a large focus on research and development of the next state-of-the-art Large Language Model.This overlooks the fact that smaller models can be run on consumer hardware.This thesis aimed to identify what aspects and attributes of a Large Language Model (LLM) with, and without Retrieval-Augmented Generation (RAG) required the most hardware.This opportunity presented itself when Alstom, a large international actor in the railroad field, wanted to look into deploying an AI Chatbot to provide support for EDAL05
Authors: Olausson, Albin, Sinclair, Ludwig
Institution: Lund University
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7067429471