Anomaly Detection

Automated Volume & Freshness Monitors

Out-of-the-box ML-powered monitoring for freshness and volume issues on all production tables. The monitors track updates to tables, and will detect data delays, incomplete updates, and significant volume changes.
By qurying only metadata (e.g. information schema, query history), the monitors don’t add compute costs.

Opt-in Anomaly Detection Monitors

ML-powered anomaly detection on data quality metrics such as null rate, empty values, string length, numeric metrics (sum, max, min, avg), etc. Elementary also supports monitoring for anomalies by dimensions. The monitors are activated for specific data sets, and require minimal configuration (e.g. timestamp column, dimensions).

Schema Validation

Schema Tests

Elementary offers a set of schema tests for validating there are no breaking changes. The tests support detecting any schema changes, only detecting changes from a configured baseline, JSON schema validation, and schema changes that break downstream exposures such as dashboards.

Automated Schema Monitors

Coming soon!

Data Tests

Custom SQL Tests

dbt tests

Python tests

Tests Coverage

Performance monitoring