UBOS Asset Marketplace: Supercharging LLMs with MCP Servers

In the rapidly evolving landscape of Artificial Intelligence, the need for Large Language Models (LLMs) to access and interact with external data sources and tools has become paramount. Enter the Model Context Protocol (MCP) Server, a pivotal component in bridging this gap. The UBOS Asset Marketplace offers a curated selection of MCP Servers, designed to enhance the capabilities of LLMs by providing them with the contextual awareness necessary for more informed and effective decision-making.

What is an MCP Server?

MCP, or Model Context Protocol, is an open standard that formalizes how applications furnish context to LLMs. An MCP Server acts as an intermediary, enabling AI models to access, process, and utilize external data. This integration is crucial because LLMs, while powerful, are limited by their training data. By connecting to external sources, MCP Servers allow LLMs to overcome these limitations and provide more relevant, accurate, and up-to-date information.

Key Features of MCP Servers on UBOS

• Data Connectivity: MCP Servers facilitate seamless connections to a wide array of data sources, including databases, APIs, and real-time data streams.
• Contextual Enrichment: By providing LLMs with access to external data, MCP Servers enrich their understanding of the context surrounding a query, leading to more accurate and insightful responses.
• Tool Integration: MCP Servers enable LLMs to interact with external tools and services, allowing them to perform actions beyond simple information retrieval.
• Customization: UBOS offers MCP Servers with customizable configurations, enabling developers to tailor their performance to specific use cases and data requirements.
• Security: Security is a top priority. MCP Servers on UBOS are designed with robust security measures to protect sensitive data and prevent unauthorized access.

Use Cases for MCP Servers

The applications of MCP Servers are vast and span across numerous industries. Here are a few compelling use cases:

• Customer Support: Integrate an LLM with a CRM database to provide customer support agents with real-time customer information, enabling them to deliver personalized and effective assistance.
• Financial Analysis: Connect an LLM to financial data feeds to provide analysts with up-to-date market information and insights, enabling them to make more informed investment decisions.
• Supply Chain Management: Integrate an LLM with supply chain data to provide managers with real-time visibility into inventory levels, delivery schedules, and potential disruptions, enabling them to optimize operations and minimize risks.
• Healthcare: Connect an LLM to patient medical records to provide doctors with quick access to relevant patient information, enabling them to make more accurate diagnoses and treatment plans.
• Knowledge Management: Use MCP Servers to connect LLMs to internal knowledge bases, enabling employees to quickly find the information they need and improve productivity.

Example: RAG API for Document-Based Q&A

Consider a Retrieval-Augmented Generation (RAG) API built on FastAPI. This API allows users to upload documents (PDFs and CSVs), store them in a vector store, and then query those documents using natural language. The MCP Server acts as the bridge between the LLM and the vector store, enabling the LLM to retrieve relevant information from the documents and generate accurate answers.

Key Functionalities of the RAG API:

1. Document Upload and Vector Store Storage:
   • Supports PDF and CSV files.
   • Automatically splits and embeds documents.
   • Stores embeddings in a Chroma vector store.
2. Document Search:
   • Enables natural language query-based search.
   • Provides similarity-based document retrieval.

Setting up the RAG API:

1. Install Dependencies:

   bash pip install -r requirements.txt

2. Configure Environment Variables:

   bash export OPENAI_API_KEY=“your-api-key”

3. Run the Server:

   bash python main.py

   The server will run at http://localhost:8000.

API Endpoints:

• Document Upload:
  • POST /upload
  • Accepts multipart/form-data with files.
  • Parameters:
    • files (required): List of files to upload.
    • vector_store_dir (optional, default: “vector_store”): Directory to store the vector store.
  • Supported formats: PDF, CSV
• Document Search:
  • POST /query
  • Accepts form-data.
  • Parameters:
    • query (required): Search query.
    • vector_store_dir (optional, default: “vector_store”): Vector store directory.
    • k (optional, default: 2): Number of documents to retrieve.

Accessing API Documentation:

• http://localhost:8000/docs
• http://localhost:8000/redoc

Reusing Vector Stores:

To reuse vector store files in another project:

1. Copy the vector store directory to the same path in the other project.
2. Ensure the following packages are installed:
   • langchain-chroma
   • langchain-openai
   • Other necessary dependencies.
3. Use the same embedding model (OpenAIEmbeddings).
4. Set the required environment variables (e.g., OpenAI API key).

Simply copying the directory preserves the document embeddings and metadata.

Using Multiple Vector Stores:

The project supports using multiple vector stores simultaneously. Each vector store is stored in a different directory, and you can specify the desired vector store using the vector_store_dir parameter in API calls.

For example:

project1_docs -> vector_store_project1 project2_docs -> vector_store_project2 research_papers -> vector_store_research

This allows you to independently manage and search each document set.

UBOS: Your Full-Stack AI Agent Development Platform

UBOS is a comprehensive AI Agent development platform designed to empower businesses across all departments. Our platform offers a suite of tools and services to orchestrate AI Agents, connect them with your enterprise data, build custom AI Agents with your LLM model, and develop Multi-Agent Systems.

Key Benefits of UBOS:

• Accelerated Development: UBOS provides pre-built components and tools to streamline the AI Agent development process, reducing time-to-market.
• Seamless Integration: UBOS seamlessly integrates with existing enterprise systems, enabling AI Agents to access and utilize valuable data.
• Enhanced Scalability: UBOS is designed to scale with your business needs, ensuring that your AI Agents can handle increasing workloads and data volumes.
• Improved Collaboration: UBOS fosters collaboration among developers, data scientists, and business users, enabling them to work together to build and deploy effective AI Agents.
• Simplified Management: UBOS provides a centralized management console for monitoring, managing, and optimizing your AI Agents.

By leveraging the power of UBOS and MCP Servers, businesses can unlock the full potential of AI and drive innovation across their organizations. Explore the UBOS Asset Marketplace today to discover the MCP Servers that best suit your needs and embark on your AI-powered transformation journey.