Vector databases have recently gained much attention for managing unstructured data in the era of big data. They efficiently store, retrieve, and analyze data in the vector form, facilitating the processing of vast amounts of data. However, challenges arise due to syntax variations among database vendors, resulting in complexities in querying and migrating between vector databases of different vendors. Additionally, the lack of a standardized language for accessing vector databases necessitates programming skills, as end-users usually access them by using programming languages, hindering user experience. The main goal of our research is to provide end-users with straightforward use of vector databases by proposing a uniform approach across various vector databases. To achieve such a goal, the development of a Model-Driven (MD) solution is proposed, with a novel Domain-Specific Language (DSL) as its core component, aiming to overcome identified challenges. As a domain analysis is a prerequisite for creating a DSL, vector database concepts and terminology are analyzed, and in this paper, the results of such a domain analysis are presented using the Feature-Oriented Domain Analysis (FODA) method, alongside a proposal for the MD solution architecture.