Artificial Intelligence (AI) has been revolutionizing various sectors with its remarkable applications. Generative AI, in particular, has gained significant traction in both industrial and research domains. In this work, we focus on leveraging this technology for an industrial use case, specifically for generating high-quality data (i.e., text data) for service request reports in manufacturing units. Addressing safety concerns is crucial in energy production sites. Field engineers raise several issues, including product safety impact, environmental concerns, health and safety, and quality issues, as well as fleet implications. The goal of our application is to classify these concerns into either unsafe or safe operating conditions, allowing the system to trigger warnings and enable engineers to prioritize and address them effectively. To implement this application successfully, a substantial amount of data is required. However, in real-world scenarios, available data is often insufficient. In this paper, we have utilized several text generation models for our application and compared their performance based on the quality of the generated text. We have successfully generated samples that closely resemble the original text using large language models. Notably, we achieved an \({\approx 30\%}\) improvement in text classification task performance by employing a large language model (GPT2, Falcon, and LLama-2).