Retrieval Augmented Generation on Hybrid Cloud: A New Architecture for Knowledge Base Systems
Chia‐Chuan Chuang, Kai-Ching Chen · 2024
Retrieval Augmented Generation (RAG) is a novel approach that combines the strengths of large language models and external documents to generate responses grounded in retrieved information. This technique is particularly useful for building knowledge base systems. However, deploying such systems requires a large amount of computational resources for hosting language models, and keep privacy of the document is also a concern. Through the nature of hybrid cloud, the knowledge base system can maintain data privacy while benefiting from the scalability of public clouds. This study leverages a hybrid cloud architecture, and utilizes Kubernetes as a cloud orchestrator to schedule the workload across different cloud environments. The implemented RAG system demonstrates the feasibility of the proposed architecture, showcasing enhanced privacy, flexibility, and security.