aegis-dq/aegis-dq

What it does

Aegis DQ is an agentic data quality framework designed to maintain data integrity across modern data warehouses. It executes structured data quality rules against multiple warehouse platforms, automatically diagnoses failures using LLM-powered root cause analysis, and generates SQL-based remediation proposals. The tool provides comprehensive audit logging for every LLM decision including costs and latency metrics.

Key Features

• Multi-warehouse support: DuckDB, BigQuery, Athena, Databricks, Postgres
• Structured rule execution against warehoused data
• LLM-powered root cause analysis for data quality failures
• Automated SQL remediation proposals
• Complete audit logging with cost and latency tracking
• MCP server implementation for integration with AI systems

How to set up

As an MCP server, Aegis DQ can be integrated into compatible AI platforms. Install via the aegisdq/aegisdq GitHub repository, configure your data warehouse credentials for the target platform (DuckDB, BigQuery, Athena, Databricks, or Postgres), define your data quality rules, and connect through the MCP protocol to your AI agent or application.