Responsible AI: Improving accuracy and reducing hallucinations in enterprise grade LLMs for B2C applications | Kisaco Research