AI-Assisted Semantic Model Generation for Prompt-to-SQL Applications

Abstract

A novel approach for automatically generating semantic models that enable large language models to accurately translate natural language queries into SQL statements. This approach addresses fundamental scalability limitations of manual semantic model and prompt creation by mining database query logs to extract usage patterns, analyzing table metadata to identify facts and dimensions, integrating data catalog information, and encoding all discovered relationships in a knowledge graph (for e.g. in Neo4j). A custom processor queries the graph to automatically generate semantic model specifications that are eighty to ninety percent complete, with human review workflows providing final validation. The approach demonstrates clear differentiation through its integration of query pattern mining with knowledge graph reasoning, enabling automatic maintenance, SQL query scoring, and cross-database semantic reconciliation beyond basic view generation.

Bios

Image

Nick Lee is lead product manager for several Data & AI initiatives at Disney Experiences, driving data strategy across Disney Parks and Products. Previously, he helped Fortune 500 companies solve complex AI problems at Databricks and Labelbox. Nick was recently recognized as a Georgia Tech Alumni Association "40 Under 40" honoree.

Image

Nandakishor Koka is a Principal Technical Architect for Data & AI at The Walt Disney Company, specializing in data engineering, machine learning, and semantic modeling at scale. He designs metadata-driven platforms, knowledge-graph-powered semantic layers, and AI systems that bridge raw data to business meaning. With a Master’s in Machine Learning from Georgia Tech, he has led innovations in computer vision, large-scale analytics, and automated semantic lifecycle management across enterprise data ecosystems.