This thesis explores the most suitable language models for improving the interaction between users and databases via natural language, reducing or completely neglecting the need for technical knowledge. We address the increasing complexity of databases and the demand for intuitive query interfaces by developing a proof-of-concept platform that transforms natural language queries into SQL-like statements for efficient data retrieval. The system supports multiple databases and enables users to analyze query results in various formats. Key contributions include benchmarking state-of-the-art models, exploring novel prompt enrichment approaches like Chain-of-Thought reasoning, designing a comprehensive system architecture, and conducting experiments to evaluate effectiveness. Our work enhances database accessibility and usability, enabling users to interact intuitively with complex data structures and improving data-driven decision-making processes.