Claude Unveils Protocol for AI-data Integration
- By Paul Mah
- November 28, 2024
Anthropic this week released a new open-source tool to connect AI assistants directly to the data sources they need to carry out tasks or inform their responses.
Specifically, the Model Context Protocol (MCP) is an open protocol that provides a universal connection to all sorts of data sources and was designed to work across all AI systems and data sources.
A powerful data connector
Developers currently have to create custom code for each dataset they want their AI model to draw from. By using MCP, developers can integrate it with their AI tool once, then leverage it for sharing resources, tools, and prompts to connect to data sources anywhere.
The power of MCP is how it handles both local resources such as databases, files, services, and remote ones such as remote API calls through the same protocol. Developers can use it to build two-way connections between data sources and AI-powered applications such as chatbots. Developers can expose data through “MCP servers” and build “MCP clients” in the form of apps and workflows that connect to those servers on command.
Anthropic says coding software like Replit, Codeium and Souregraph have already started using MCP to build out their AI agents. The tool will make it easier for other companies and developers to connect an AI system with multiple data sources.
“Instead of maintaining separate connectors for each data source, developers can now build against a standard protocol. As the ecosystem matures, AI systems will maintain context as they move between different tools and datasets, replacing today's fragmented integrations with a more sustainable architecture,” said Anthropic in its announcement.
While MCP sounds like a great idea, adoption remains unclear, especially as rivals such as OpenAI would likely prefer to use their own proprietary approaches and specifications, as well as build up its own ecosystem. It is also unclear whether MCP will be as beneficial or perform as well as Anthropic claims.
For now, developers can start building and testing MCP connectors today by installing pre-built MCP servers through the Claude Desktop app. You can read more about Anthropic’s MCP here.
Image credit: iStock/Vitalii Gulenok
Paul Mah
Paul Mah is the editor of DSAITrends, where he report on the latest developments in data science and AI. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose.