MetaPatch

MetaPatch is a web app (with optional CLI) that continuously audits metadata quality across your warehouse/lakehouse and BI layer, then opens actionable “patches” to fix it. It connects to Snowflake/BigQuery/Databricks plus dbt and key BI tools, scans for common metadata failures (missing owners, stale descriptions, inconsistent tags, PII fields unclassified, broken lineage links, orphaned tables, undocumented metrics), and produces a prioritized backlog. The AI layer suggests concrete fixes: draft column/table descriptions, propose standardized tag sets, infer owners from git/queries, and generate pull requests for dbt YAML or catalog APIs—while keeping humans in approval. It’s not a full data catalog replacement; it’s a metadata maintenance system that sits on top of what you already have. Expect some setup and tuning, but the goal is fewer “tribal knowledge” incidents and faster onboarding for analysts and engineers.