SchemaSentry

SchemaSentry monitors metadata “drift” across your warehouse, BI layer, and ETL jobs, then alerts you when definitions, column meanings, or ownership silently change. It connects to common stacks (Snowflake/BigQuery, dbt, Looker/Tableau) and builds a lightweight catalog focused on what actually breaks: renamed fields, type changes, duplicated metrics, and undocumented transformations. The app continuously compares current schemas and semantic models to a baseline, flags risky changes, and generates a human-readable changelog for data consumers. It also enforces simple governance rules (e.g., PII tags required, owners assigned, freshness thresholds) and routes issues to Slack/Jira with suggested fixes. This is not a “giant catalog replacement”; it’s a pragmatic guardrail that teams can adopt without a months-long rollout. Web app with optional CLI for CI checks; AI-assisted explanations and auto-tag suggestions.