As artificial intelligence emerges as a driving force for future development, its integration with legacy systems offers unprecedented opportunities for modernization and operational efficiency. This study introduces a novel framework that bridges the gap between generative AI models and enterprise-specific structured knowledge. By leveraging pre-retrieval mechanisms, the framework dynamically integrates relevant data from legacy systems into AI responses, ensuring contextual relevance and precision. The proposed approach enhances usability by combining organizational data with generative AI, making it particularly effective for internal knowledge management and innovation. Two educational case studies - course recommendation and self-directed learning - demonstrate the framework’s applicability in real-world scenarios. The results highlight its ability to provide accurate, context-aware responses while preserving system stability and scalability. This approach not only addresses the limitations of pre-trained models but also establishes a practical pathway for sustainable AI adoption in enterprise environments.