The following study was completed as part of the requirements for the Master of Science in Data Science. This paper was a case study to fine tune LLMs on retail domain data. The results of this study will provide a retail associate with the capability to make simple requests for product information and customer sentiment from a wrist worn device to assist in-store customers with queries to improve in-store purchase intent. Since LLMs are initially trained for a general use, performance can be improved by tuning them to the specific domain where they will be used by using data specifically from the retail domain in this case. This study fine-tuned LLMs using PEFT methods to optimize classification and summarization results on the retail domain dataset. Retail associates can then use their wrist devices with their common retail lexicon to request product data or customer sentiment and be confident the inference results will be high quality.