Report Abuse

Basic Information

Wren Engine is a semantic engine designed to enable MCP clients and AI agents to access enterprise data with context, accuracy, and governance. The repository provides the core components to interpret natural language intent, map it to the correct data model and calculations, and serve semantic-aware queries and responses for agentic workflows. It is built to be embeddable into MCP clients and interoperable with modern data stacks so LLM-driven agents can perform trusted reporting, lookups, and updates across warehouses, databases, and file stores. The README highlights use at the enterprise level where precision, business term clarity, and user-based permissions are required. The project includes server and core modules, mentions semantic SQL concepts and a modeling definition language (MDL), and notes that the Wren AI GenBI agent is based on Wren Engine.

Links

Categorization

App Details

Features
The repo bundles multiple modules: a FastAPI and Ibis powered web server (ibis-server), a Rust semantic core (wren-core) leveraging Apache DataFusion, Python bindings (wren-core-py), and an MCP server implementation (mcp-server) using the MCP Python SDK. It provides connectors for many data sources listed in the README including BigQuery, Google Cloud Storage, local files, MS SQL Server, Minio, MySQL, Oracle, PostgreSQL, Amazon S3, Snowflake, and Trino. The engine emphasizes semantic-first handling, semantic SQL generation, a Modeling Definition Language for formalizing business terms, embeddability into MCP clients, interoperability with enterprise data stacks, and governance features such as role-aware access control. The project is noted as beta and actively maintained.
Use Cases
Wren Engine helps organizations run AI-driven workflows that require precise, auditable access to structured enterprise data. It translates agent intent into semantically correct queries and ensures calculations, aggregations, and business definitions (for example, "active customer" or "net revenue") are applied consistently. By embedding the semantic layer into MCP clients, it reduces errors from raw data access, enforces access controls and governance policies, and makes AI interactions explainable and repeatable across BI, CRM, and compliance workflows. The supported data connectors and language bindings enable teams to integrate the engine into existing stacks and MCP-enabled clients so agents can act on data with context-aware correctness rather than relying solely on raw natural language or direct database queries.

Please fill the required fields*