Data Observability

Use the Datachecks Data Observability dashboard in Datachecks provides real-time insights into the health and reliability of your data

Data Health Overview

Health Score:

Displays the overall quality score of your data. A declining trend indicates deteriorating data health and signals potential issues that may need investigation

Schema Coverage:

Tracks how much of your data's schema is covered by validations. This helps ensure your data structure remains consistent and accurate over time.

Data Health:

The Data Health graph visualizes the overall quality of your data over a selected time frame. A stable or increasing trend indicates good data quality, while a sharp drop may suggest anomalies or failures. When hovered over, it displays the detected time range and the overall health score trend within that period.

Validation Health Filter

Here, you can view the types of validations, the total number of validations of each type within the current workspace, and their respective health scores. By selecting a specific validation type, its trend will be reflected on the Data Health graph. Hovering over the graph displays the detected time range, the overall health score, and the selected validation score trend within that period.

Datasets Overview

Below the health score, the table provides a detailed overview of the validation results for your datasets

FieldDescriptionStatus
SummaryA quick overview of the validation status of your datasets.Pass / Fail
DatasetDisplays the dataset name to help you identify and monitor data sources.Pass / Fail
Pipeline ReliabilityIndicates the consistency and reliability of data pipelines.Pass / Fail
UniquenessEnsures no duplicate data exists within the dataset.Pass / Fail
CompletenessTracks the presence of missing data points.Pass / Fail
DistributionsMonitors the distribution of data to detect any abnormalities.Pass / Fail

When you collapse a particular dataset, you'll access the Historical Source Scorecard. This view displays the validation types applied to the dataset and their overall health on specific dates, represented using color codes. Hover over a health status to view the detected time range.