TroveGrid
Back to all tools
Soda logo

Soda

YAML-defined data quality checks — monitor, alert, and resolve data issues at scale.

0Free
Data Quality
Official Website View on GitHub
Share:XLinkedInWhatsApp

Overview

Soda is a data quality platform that uses a YAML-based check language (SodaCL) to define, run, and monitor data quality rules across SQL databases, data warehouses, and data pipelines with built-in alerting and collaboration.

Key Features

  • SodaCL — a human-readable YAML check language for data quality rules
  • Anomaly detection checks powered by ML for automatic threshold learning
  • Native integrations with Snowflake, BigQuery, Redshift, Spark, and DuckDB
  • Soda Cloud for collaborative incident management and check history
  • Programmatic checks via the Soda Python library
  • dbt-soda integration to run checks on dbt model outputs

Real-World Workflows

Monitor data freshness and completeness

  1. 1Write a SodaCL checks.yml with freshness and row count checks
  2. 2Run soda scan on a schedule or in the CI/CD pipeline
  3. 3Alert on Slack or PagerDuty when checks fail
  4. 4Track check history and trends in Soda Cloud

Integrate data quality into dbt pipelines

  1. 1Configure soda-core-dbt adapter in your dbt project
  2. 2Add SodaCL checks for critical dbt model outputs
  3. 3Run soda scan after dbt run in your Airflow or Dagster DAG
  4. 4Gate downstream pipeline steps on data quality pass/fail status

Getting Started

# Install Soda Core with Snowflake
pip install soda-core-snowflake

# Create a configuration file
# configuration.yml:
# data_source my_warehouse:
#   type: snowflake
#   account: ...

# Create a checks file
# checks.yml:
# checks for orders:
#   - row_count > 0
#   - missing_count(status) = 0

soda scan -d my_warehouse -c configuration.yml checks.yml

Compare Alternatives

See how Soda stacks up against similar tools.

VSgreat expectationsVSdeequVSre dataVSmonte carlo