Generative AI, specifically retrieval-augmented generation, is transforming life sciences and healthcare compliances compliances by improving decision-making processes. AI systems may provide specialized reports, projections, and investment plans by analyzing massive volumes of pharmaceutical and contextual information. These systems use retrieval techniques to generate accurate and contextually relevant outputs based on historical data, market trends, and expert views. This work describes a novel evolutionary strategy for improving connection awareness in retrieval augmented generation (RAG) systems, with an emphasis on compliance documents for life sciences and healthcare compliances industry. Using the power of knowledge graphs and graph databases, our approach provides a comprehensive framework for modeling documents and their deep relationships, allowing for more effective information retrieval and creation processes. We offer a method for leveraging retrieval-augmented generation from a graph database, in which documents are represented as nodes and relationships as edges, allowing for the extraction of rich contextual information. Furthermore, we present a similarity-based searching strategy for the graph database, allowing for more precise and relevant document retrieval. To assess the efficacy of our technique, we undertake a life sciences and healthcare compliances compliances case study that examines the effects of relationship-aware retrieval enhanced generation on important metrics including ROUGE and BLEU. We show considerable gains in these indicators after iterative experimentation, demonstrating that the created information is of higher quality and relevance. By including relationship awareness into the retrieval augmented generation process, our method allows finance professionals to access and develop insights with improved clarity, accuracy, and contextually. Our findings indicate the potential for using knowledge graphs and graph databases to improve the capabilities of retrieval-augmented generation systems in life sciences and healthcare compliances compliances applications. Beyond typical keyword-based retrieval approaches, our approach provides a more complex knowledge of document linkages, resulting in better decision-making processes. Furthermore, the iterative nature of our evolutionary approach enables continual refinement and adaptation, guaranteeing that the system stays effective in dynamic and changing situations. Finally, our findings add to ongoing efforts to advance retrieval augmented generation techniques by emphasizing the importance of relationship awareness and the use of knowledge graphs. By incorporating these concepts into the fabric of life sciences and healthcare compliances applications, we open the door to more intelligent, context-aware systems that provide finance professionals with actionable insights and decision support