AI Connectivity via MCP

Learn how BlueNexus connects user data to your AI agent or application

BlueNexus AI Connectivity lets you bring any user‑owned data into a large‑language‑model (LLM) workflow without moving or duplicating the data. The bridge is the Model Context Protocol (MCP) – a thin, secure service layer that is provided by BlueNexus for every third‑party data source a user connects (Health data, CRM, analytics platform, etc.).

  • One MCP server per data source – each source gets its own isolated endpoint that knows how to translate LLM‑driven intents into the native query language of that source.

  • Unified programming model – the same HTTP‑based protocol works for Mongo‑style collections, SaaS APIs, flat‑file stores, or any custom connector you register.

The result is an AI‑enabled application that can ask questions like:

  • “What were the total sales for the top‑3 products in the last 30 days?”

  • "What was my average resting heart‑rate during the past 7 days, and did I have any day where it spiked above 100 bpm?"

Try it now with the Connect User Data Example

How it works

  1. The LLM will analyse the user prompt and identify the MCP tool call that should be executed

    1. "What was my average resting heart‑rate during the past 7 days"

    2. Tool Identified: BlueNexus_HeartRate

  2. The LLM will call the BlueNexus MCP tool

    1. Call Tool: BlueNexus_HeartRate(startDate, endDate, "HeartRateAvg,HeartRateMax,HeartRateMin")

    2. Sends API request to BlueNexus MCP server

    3. BlueNexus MCP Server locates user credentials for the heart rate data source

    4. BlueNexus MCP Server submits authenticated request to third party API

    5. BlueNexus MCP Server receives response, standardizes the response format

    6. BlueNexus MCP Server sends response back to LLM

  3. The LLM receives the MCP Server response and uses that to product a human-readable answer.

Further reading

  1. Model Context Protocol Documentation

PreviousOverviewNextConnect User Data Example

Last updated