Construct SQL from table records by breaking down decision tree

This project automatically generates SQL queries from tabular data by overfitting a decision tree to the records, then simplifying the resulting boolean logic into concise, human-readable SQL conditions. It is designed for data analysts, ML engineers, and developers who need to quickly reverse-engineer query logic from example data or debug classification rules. The approach is interesting because it flips traditional machine learning—using overfitting not as a flaw but as a tool for transparent, rule-based SQL generation, with a live demo on Streamlit.