LLM fine tuning for E-commerce product recommendation

Hello, I’m in the process of developing an Ecommerce chatbot that offers product recommendations tailored to user input.
I’m interested in training the model to comprehend product details such as price, release date, model number, summary, and purchase links.
Additionally, I’d like to ensure that this product information remains up-to-date. Is it feasible to establish a connection with a data source like an SQL database to achieve this?
Resource link would be helpful.