From Data Contracts to Talking to your Data, with Entropy Data

Entropy Data addresses a fundamental problem with traditional data catalogs that attempt to automatically index everything in a data warehouse, creating thousands of unusable entries where finding authoritative data sources becomes impossible. Their solution centers on manually curated data contracts that serve as concrete YAML specifications defining ownership, schema, semantics, governance rules, quality metrics and explicit terms of use for data products.

The company offers both an open-source Data Contract CLI for validating that implementations match specifications and a commercial platform functioning as a data product marketplace. Available in SaaS, the platform supports major data platforms including Snowflake, Databricks, Postgres and BigQuery. Entropy Data is actively contributing to industry standardization through the Open Data Contract Standard under the Linux Foundation's Bitol project and plan to join the Open Semantics Interchange initiative alongside Snowflake, dbtLabs, and Tableau.

The most compelling innovation is the platform's AI integration through the Model Context Protocol. By feeding rich metadata from data contracts into LLM context, users can query datasets in natural language while the system constructs accurate SQL queries. The data contract acts as an active governance layer, not just documentation, enforcing rules at both client and server side to prevent unauthorized actions like exporting customer data for prohibited marketing purposes and protecting against security risks like prompt injection. This transforms metadata from passive documentation into an enforceable interface enabling intelligent, secure data access.