In a world increasingly shaped by AI, data is the critical fuel that powers everything — making seamless connectivity across silos more essential than ever.
The OpenLink AI Layer (OPAL) represents a fundamental leap in intelligent data integration—one that collapses the complexity of AI agent creation into a streamlined, standards-based, and production-ready approach.
In this article, we showcase how OPAL enables the rapid creation and deployment of AI agents that emulate thinking in context and communicate using natural language—all backed by robust connectivity to your enterprise data spaces.
What is OPAL?
OPAL (OpenLink AI Layer) is a lightweight yet powerful framework for building AI agents that can interface with enterprise data systems, query knowledge graphs, and interact using natural language—all without the usual integration overhead.
What Makes OPAL Different?
Traditionally, building a useful AI agent meant stitching together four separate workflows:
- Conceptualization
- Coding
- Testing
- Documentation
With OPAL, these become one.
OPAL-based agents are authored using a single artifact—written in Markdown or JSON—which fuses these traditionally distinct tasks into a unified definition. That same file serves as:
- A design document
- A functioning prototype
- A test bench
- A deployable runtime
This streamlined approach enables faster iteration, better transparency, and easier maintenance.
Enterprise-Ready by Design
Behind the scenes, OPAL taps into the rich capabilities of OpenLink’s Data Access Layer, making AI agents accessible via:
- ODBC/JDBC, courtesy of the Model Context Protocol (MCP)
- RESTful APIs
- Direct SPARQL endpoints and knowledge graphs
OPAL ships with its own native MCP Server, enabling agents to be treated as virtual data sources—queryable like traditional databases from familiar tools.
Explore the OPAL Installation & Usage Guide »
Context is the Superpower: Enter MCP
At the heart of OPAL’s intelligence lies the Model Context Protocol (MCP)—a semantic bridge between models, data, and context-aware computing.
With MCP, OPAL agents gain contextual intelligence. They don’t just access data—they understand what that data means in relation to your information model, ontologies, and business logic.
OPAL in the Real World: Cello, the EUPO Virtual Assistant
To see OPAL in action, meet Cello, a virtual assistant created for the EU Publications Office (EUPO). Cello leverages OPAL and MCP to navigate complex legislation data, extract rich metadata, and deliver natural language responses.
Demo 1: Navigating Publications & WEMI Structures
From request to response, Cello’s interaction flow shows OPAL’s ability to bridge language and data.
1. Find Publications
“Please find publications about renewable energy”
Watch Demo
2. Get Business Identifiers & Titles
3. Extract WEMI Hierarchies
“Extract a WEMI hierarchy from CELEX 32024R1106”
Demo 2: Ontology-Aware Reasoning
This second demo illustrates OPAL’s capacity to generate and interpret SPARQL queries over a CDM ontology—without hardcoded logic.
1. Activate CDM Ontology
“I’m looking for information about secondary legislation”
2. Search Legislation Relationships
“Find secondary legislation that modifies another legislation”
3. Query Trace & Transparency
“What query did you run?”
Why This Matters
For Existing Customers & Partners
OPAL is an immediate upgrade to your current infrastructure—bridging natural language interfaces with your structured databases, knowledge graphs, and ODBC-accessible systems. With its built-in MCP Server, OPAL agents are not only conversational—they’re also queryable through BI tools, APIs, and data fabrics.
For Future Customers & Partners
OPAL offers a shortcut to practical, enterprise-ready AI—without the learning curve of massive LLM infrastructure or agent orchestration platforms. If your team knows Markdown, JSON, and SQL (or SPARQL), you’re already ready.
“Data Silos are Killing Your Data Flow”
Get Started
OPAL Installation & Usage Guide
Explore the UniProt Assistant
See Cello’s Configuration
Watch the OPAL Overview Presentation
Final Thought
With OPAL, the future of AI agent development is already here: unified, contextual, and deployable across your enterprise data fabric. Say goodbye to fragmented workflows and disconnected systems. Say hello to intelligent agents that work the way you think.