Pipeline Auditor

Pipeline Auditor is a web and desktop app focused on one brutal reality: most data teams have no idea when their pipelines silently start rotting. Instead of trying to be a full-blown data platform, it does one thing well—detects, explains, and helps fix data pipeline issues across warehouses and ETL tools. It connects read-only to sources like Snowflake, BigQuery, Redshift, dbt, and Airflow, then builds a dependency graph of tables, jobs, and dashboards. Using anomaly detection and rule-based checks, it flags schema drifts, broken joins, unexpected null spikes, volume drops, and freshness violations. The AI layer suggests likely root causes in plain language and proposes concrete fixes (e.g., “add this column mapping in dbt model X” or “update this Airflow DAG schedule”). Instead of dashboards nobody checks, it pushes prioritized alerts into Slack, email, and Jira with clear impact estimates (e.g., “affects 3 exec dashboards and 2 ML models”). It’s opinionated, slightly ruthless, and assumes your data stack is already a mess—because it usually is.