UBOS Asset Marketplace: CData MCP Server for Google BigQuery - Unleash the Power of AI on Your Data

In today’s data-driven world, the ability to quickly access and analyze information is paramount. Large Language Models (LLMs) like Claude are revolutionizing how we interact with data, enabling us to ask natural language questions and receive insightful answers. However, LLMs need a bridge to connect to the vast sea of data stored in various databases and applications. This is where the Model Context Protocol (MCP) and UBOS come into play.

The UBOS Asset Marketplace offers a comprehensive solution: the CData MCP Server for Google BigQuery. This read-only MCP Server empowers you to connect Claude Desktop to your Google BigQuery data using CData JDBC Drivers, unlocking the potential of AI-driven data access. For those seeking read/write capabilities, CData offers a free (beta) version of their MCP Server. This integration seamlessly connects your Google BigQuery data to LLMs, enabling natural language queries and real-time insights.

Understanding MCP: The Key to Contextual AI

Before diving into the specifics of the CData MCP Server, let’s understand the foundational technology: Model Context Protocol (MCP). MCP is an open protocol designed to standardize how applications provide context to LLMs. Think of it as a universal translator, allowing AI models to understand and interact with different data sources and tools.

Why is MCP important?

• Bridging the Gap: MCP bridges the gap between LLMs and the external world. It allows AI models to access and process data from various sources, providing them with the necessary context to answer complex questions.
• Standardized Communication: MCP establishes a standardized way for applications to communicate with LLMs, ensuring consistency and compatibility.
• Enhanced AI Capabilities: By providing access to real-time data and external tools, MCP enhances the capabilities of AI models, enabling them to perform more sophisticated tasks.

CData MCP Server for Google BigQuery: A Deep Dive

The CData MCP Server for Google BigQuery acts as a specialized bridge, connecting your Google BigQuery data to LLMs like Claude. It leverages the CData JDBC Driver for Google BigQuery, which exposes Google BigQuery as relational SQL models. This allows the MCP Server to translate natural language queries into SQL queries, retrieve the relevant data, and present it to the LLM in a format it can understand.

Use Cases:

• Real-Time Data Analysis: Ask Claude questions about your Google BigQuery data and receive real-time answers. For example, “What is the total revenue generated by product X in the last quarter?”
• Data-Driven Decision Making: Use Claude to analyze trends in your data and identify opportunities for improvement. For example, “What are the key factors driving customer churn?”
• Automated Reporting: Automate the creation of reports by asking Claude to generate summaries of your data. For example, “Create a report summarizing the sales performance for the last month.”
• Enhanced Customer Service: Integrate Claude with your customer service platform to provide agents with real-time access to customer data stored in Google BigQuery. For example, “What is the customer’s purchase history?”
• Improved Business Intelligence: Gain deeper insights into your business by combining the power of LLMs with your Google BigQuery data.

Key Features:

• Read-Only Access: This version of the MCP Server provides read-only access to your Google BigQuery data, ensuring data integrity.
• CData JDBC Driver Integration: Leverages the robust and reliable CData JDBC Driver for Google BigQuery.
• Natural Language Queries: Enables you to query your data using natural language, eliminating the need to write complex SQL queries.
• Real-Time Data Access: Provides real-time access to your Google BigQuery data, ensuring you always have the latest information.
• Easy Setup: The server is easy to set up and configure, allowing you to quickly connect your data to LLMs.
• MCP Compliance: Adheres to the Model Context Protocol (MCP) standard, ensuring compatibility with various LLMs.

Setting Up the CData MCP Server for Google BigQuery

The setup process involves several key steps:

1. Clone the Repository: Begin by cloning the GitHub repository containing the MCP server code:

   bash git clone https://github.com/cdatasoftware/google-bigquery-mcp-server-by-cdata.git cd google-bigquery-mcp-server-by-cdata

2. Build the Server: Use Maven to build the server, creating a JAR file with all necessary dependencies:

   bash mvn clean install

   This generates CDataMCP-jar-with-dependencies.jar.

3. Install the CData JDBC Driver: Download and install the CData JDBC Driver for Google BigQuery from the CData website.

4. License the Driver: License the CData JDBC Driver by running the command java -jar cdata.jdbc.googlebigquery.jar --license in the driver’s lib folder and entering your licensing details.

5. Configure the Connection: Use the Connection String utility (java -jar cdata.jdbc.googlebigquery.jar) to configure the connection string and test the connection to your Google BigQuery instance. Copy the connection string for the next step.

6. Create a .prp File: Create a .prp file (e.g., google-bigquery.prp) to store the connection properties. This file should include the following properties:

   env Prefix=googlebigquery ServerName=CDataGoogleBigQuery ServerVersion=1.0 DriverPath=PATHTOcdata.jdbc.googlebigquery.jar DriverClass=cdata.jdbc.googlebigquery.GoogleBigQueryDriver JdbcUrl=jdbc:googlebigquery:InitiateOAuth=GETANDREFRESH; Tables=

7. Configure Claude Desktop (Example): Add the new MCP server to the Claude Desktop configuration file (claude_desktop_config.json). The path to this file will vary depending on your operating system.

   Windows Example:

   { “mcpServers”: { “{classname_dash}”: { “command”: “PATHTOjava.exe”, “args”: [ “-jar”, “PATHTOCDataMCP-jar-with-dependencies.jar”, “PATHTOgoogle-bigquery.prp” ] } } }

   Linux/Mac Example:

   { “mcpServers”: { “{classname_dash}”: { “command”: “/PATH/TO/java”, “args”: [ “-jar”, “/PATH/TO/CDataMCP-jar-with-dependencies.jar”, “/PATH/TO/google-bigquery.prp” ] } } }

8. Run/Refresh Client: Restart your Claude Desktop client to recognize the new MCP server.

Running the Server Manually

You can also run the MCP server independently using the following command:

bash java -jar /PATH/TO/CDataMCP-jar-with-dependencies.jar /PATH/TO/Salesforce.prp

Note that this server uses stdio and can only be used with clients running on the same machine.

How UBOS Enhances the MCP Server Experience

While the CData MCP Server provides the crucial bridge between LLMs and your Google BigQuery data, the UBOS platform elevates the entire experience. UBOS is a full-stack AI Agent Development Platform designed to empower businesses with AI agents across all departments.

Benefits of Using UBOS with the CData MCP Server:

• AI Agent Orchestration: UBOS allows you to orchestrate multiple AI agents, creating sophisticated workflows that leverage the CData MCP Server for data access.
• Enterprise Data Connectivity: UBOS seamlessly connects your AI agents with your enterprise data, including data accessed through the CData MCP Server.
• Custom AI Agent Development: UBOS provides the tools and infrastructure you need to build custom AI agents that leverage the CData MCP Server to meet your specific business needs.
• LLM Model Integration: UBOS allows you to integrate your own LLM models with the CData MCP Server, giving you complete control over the AI processing.
• Multi-Agent Systems: UBOS enables you to build multi-agent systems that coordinate and collaborate to solve complex problems, leveraging the CData MCP Server for data access and analysis.

UBOS empowers you to:

• Build Custom AI Agents: Develop agents tailored to your specific business requirements, enabling them to interact with your Google BigQuery data through the CData MCP Server.
• Orchestrate AI Workflows: Create complex AI workflows that automate tasks and processes, leveraging the CData MCP Server for data retrieval and analysis.
• Connect to Enterprise Data: Seamlessly connect your AI agents to your Google BigQuery data and other enterprise data sources, providing them with the context they need to make informed decisions.
• Deploy and Manage AI Agents: Easily deploy and manage your AI agents, ensuring they are always available and performing optimally.

Tools & Descriptions

Once the MCP Server is configured, the AI client can use built-in tools to interact with the data. Here are some available tools:

• {servername}_get_tables: Retrieves a list of tables available in the data source.
• {servername}_get_columns: Retrieves a list of columns for a table.
• {servername}_run_query: Executes a SQL SELECT query.

JSON-RPC Request Examples

For scripting requests to the MCP Server (without using an AI Client), refer to the JSON payload examples below:

google_bigquery_get_tables

{ “jsonrpc”: “2.0”, “id”: 1, “method”: “tools/call”, “params”: { “name”: “google_bigquery_get_tables”, “arguments”: {} } }

google_bigquery_get_columns

{ “jsonrpc”: “2.0”, “id”: 2, “method”: “tools/call”, “params”: { “name”: “google_bigquery_get_columns”, “arguments”: { “table”: “Account” } } }

google_bigquery_run_query

{ “jsonrpc”: “2.0”, “id”: 3, “method”: “tools/call”, “params”: { “name”: “google_bigquery_run_query”, “arguments”: { “sql”: “SELECT * FROM [Account] WHERE [IsDeleted] = true” } } }

Troubleshooting

1. Server Not Visible: Ensure you’ve fully quit and restarted the Claude Desktop client.
2. Data Retrieval Issues: Double-check the connection configuration and connection string in the .prp file.
3. Connection Problems: Contact the CData Support Team.
4. MCP Server Issues: Join the CData Community.

License

This MCP server is licensed under the MIT License.

Conclusion

The CData MCP Server for Google BigQuery, combined with the power of the UBOS platform, offers a transformative solution for accessing and analyzing your data with AI. By bridging the gap between LLMs and your Google BigQuery data, this integration unlocks new possibilities for real-time data analysis, data-driven decision making, automated reporting, and improved business intelligence. Embrace the future of data interaction with the CData MCP Server and UBOS.