# Zero Thresholds, Full Anomaly Detection: AWS Glue Data Quality Now Watches Your Data For You!

Hi, it's Shiichan! I found another AWS Glue Data Quality update to share with you today.

## What was announced?

According to AWS What's New, AWS Glue Data Quality now supports anomaly detection and the ability to write evaluation results to AWS Glue Data Catalog (GDC) tables. Both capabilities work across ETL jobs and Catalog-based evaluations, so you get a consistent data quality experience no matter which workflow you use.

## The story so far

Until now, spotting unexpected changes in your data meant writing explicit threshold rules yourself. But changes like a sudden drop in distinct values or a spike in row count aren't always easy to define a fixed threshold for. For data engineers monitoring hundreds of tables in GDC, that was a lot of manual work.

## What changes

The new anomaly detection uses ML-powered time-series forecasting to flag unexpected changes in your GDC table statistics, without you having to write threshold rules at all. Things like a sudden drop in distinct values or a spike in row count get surfaced automatically.

On top of that, Data Quality rule outcomes, profiling metrics, and anomaly predictions (with confidence bounds) are now written back to GDC tables, creating a queryable record of every quality evaluation. Whether it runs as part of an ETL job or directly on a Catalog table, you can query the results anytime with standard SQL.

## Wrap-up

- AWS Glue Data Quality now supports anomaly detection and writing results to AWS Glue Data Catalog (GDC)
- Anomaly detection uses ML-based time-series forecasting, no manual thresholds required
- Rule outcomes, profiling metrics, and anomaly predictions (with confidence bounds) land in GDC tables, queryable via SQL
- Works for both ETL jobs and Catalog evaluations, available in all AWS commercial Regions and AWS GovCloud (US)

If you're tired of hand-managing thresholds across hundreds of tables, this update is for you!