At AtScale, we’ve always believed that the semantic layer is the cornerstone of enterprise data strategy. It brings consistency, governance, and business context to analytics. But as generative AI enters the enterprise at full speed, the stakes and the opportunities have changed.
Model Context Protocol (MCP), a new open standard for AI-native data access, is one of the most transformative technologies we’ve seen. Let’s take a deep dive into this new standard and focus on how MCP can leverage a semantic layer to make AI and agentic AI work for the enterprise.
What Is MCP and Why Does It Matter?
The Model Context Protocol (MCP) is an open protocol that enables AI agents and LLMs to securely access metadata and data in other platforms, including governed semantic models.
With MCP, any LLM or intelligent agent can:
- Query semantically-rich data models
- Understand business definitions, hierarchies, and metrics
- Enforce role-based data access
- Operate without duplicating or copying data
In short, MCP extends the value of the semantic layer from BI tools to any AI application or agent. We implemented the AtScale MCP Server as a lightweight, containerized service, which can be deployed with minimal friction. And because it’s open, it’s designed to interoperate with any chatbot or AI agent that speaks the protocol.
Enterprise AI needs to be open, extensible, and secure. The days of vendor lock-in and siloed data architectures are numbered. MCP is modeled after protocols like JDBC: it’s not proprietary, and that’s intentional.
We’re already seeing interest from partners and customers alike. This is a tide that lifts all ships. However, what differentiates AtScale is the depth of our metadata, the power of our semantic engine, coupled with robust, built-in data governance. All this comes in a package with almost 13 years of enterprise experience and maturity.
From BI to AI: A Paradigm Shift in Enterprise Data Access
Over the last decade, enterprises have invested in data platforms, warehouses, and BI tools to democratize data access. But natural language interfaces and LLMs are shifting the expectation: users want to go beyond reports and dashboards and converse with their data.
The problem? LLMs can’t do much with databases and data lakes with hundreds or thousands of raw tables. They need business context. They need governance. They need to deliver accurate and consistent answers.
We’ve seen tools like Databricks Genie and Snowflake Cortex Analyst try to bridge this gap. But they’re often limited in scope, tightly coupled to their platforms, and unable to deliver rich, governed business context at scale.
We built the AtScale MCP Server to deliver on the promise of NLQ and agentic AI by leveraging our deep expertise in delivering semantic layer platforms to the enterprise.
One Model, Many Interfaces: The New Standard for Data Architecture
With AtScale and MCP, we’re not just adding an AI endpoint. We’re building a new abstraction layer for enterprise data access:
- BI tools speak to AtScale through SQL and MDX
- AI tools speak to AtScale through MCP
- Data scientists can use Python and REST
It’s one model, defined once, consumed everywhere. That’s what a true universal semantic layer looks like.
Building the Future, Not Chasing It
AtScale has always focused on helping enterprises make data more accessible, more consistent, and more trusted. MCP is a natural evolution of that mission. By enabling open, governed AI access through the semantic layer, we’re not just supporting chatbots; we’re empowering the next generation of enterprise intelligence.
If you’re exploring how to make your data AI-ready without losing control, it’s time to look at MCP.
Learn more about AtScale’s semantic layer and MCP protocol in this interactive demo.
SHARE
Case Study: Vodafone Portugal Modernizes Data Analytics