AI-assisted building knowledge extraction and standardization pipeline for energy analysis

This repository provides an AI-assisted pipeline for extracting and standardizing building knowledge from heterogeneous and unstructured documents for energy analysis. The workflow integrates document parsing, information extraction, and schema-based normalization to produce structured, machine-readable representations of building systems, including metadata, envelope, HVAC, energy systems, and measurement data.

The structured outputs enable consistent interpretation of building information and support data-driven energy analysis. The repository includes the full implementation, example data, and generated outputs to ensure reproducibility and facilitate reuse in research and practical applications.