Glean Drawing Index Pipeline

Work · primary developer

A proof of concept that extracts drawing metadata with OCR so the sheets can be indexed for search.

What I built

OCR extracts metadata from each sheet.
Results are written to Delta tables to detect changes between runs.
Notebooks run in Microsoft Fabric, with the code packaged as wheel files.

Stack

PythonOCRDelta LakeMicrosoft FabricPython wheels