Glyph: A Multi-Strategy Agentic System for Column Description and Sensitivity-Ontology Tagging of Enterprise Data Catalogs

What moved

Enterprise data lakes accumulate tables faster than human stewards can document or classify them, leaving columns with missing descriptions and unassigned governance labels. This documentation debt undermines data discovery, access control, and regulatory compliance. We present Glyph, a production system that frames two coupled problems, column description generation and column type annotation for data classification, as cooperating LLM agents orchestrated as stateful graphs. The Descriptor grounds generation in the pipeline source code that produces each column, retrieved on demand from.

Why it matters

Enterprise data lakes accumulate tables faster than human stewards can document or classify them, leaving columns with missing descriptions and unassigned governance labels. That is a public research file from Apple, dated 2026-09-16. Tagged LLM / Agents.

On the record

• Filed from the Apple official RSS on 2026-09-16.
• Primary source host: machinelearning.apple.com.
• Enterprise data lakes accumulate tables faster than human stewards can document or classify them, leaving columns with missing descriptions and unassigned governance labels.