Add Model Context Protocol (MCP) Client Support in Oracle Analytics (AI Assistant & Connections) — Oracle Analytics

Oracle Analytics Cloud and Server Idea Lab

Welcome to the Oracle Analytics Community: Please complete your User Profile and upload your Profile Picture

Add Model Context Protocol (MCP) Client Support in Oracle Analytics (AI Assistant & Connections)

Needs Votes
55
Views
2
Comments
Joel
Joel Rank 8 - Analytics Strategist

Idea Summary

Enable Oracle Analytics to act as a Model Context Protocol (MCP) client, allowing it to connect to MCP servers for contextual, domain-specific AI augmentation. This could be implemented either as (or both):

  • A new connection type in OAC, or
  • A feature enhancement to the built-in AI Assistant, letting it consume and query MCP servers directly.

Proposed Solution

  • Native MCP Client Connector – Add an option in OAC’s connection library for MCP, with fields for server URL, authentication method, and configuration settings.
  • AI Assistant Enhancement – Allow the AI Assistant to register MCP “tools” and “resources” dynamically, pulling in context-aware commands from external MCP servers without leaving OAC.
  • Security & Governance – Leverage OAC’s existing OCI IAM integration to secure MCP calls and respect data access policies.

Feasibility

  • MCP is an open protocol and can be implemented over standard web technologies.
  • OAC already supports custom connections and AI Assistant integration with external services, making MCP client support a logical extension.
2
2 votes

Needs Votes · Last Updated

Comments

  • Hi Joel - thanks for your suggestion,

    What sort of specifc use case do you have in mind? Are there any current MCP servers that you think would be useful to be able to call from OAC?

    Thanks

    Mike

  • Joel
    Joel Rank 8 - Analytics Strategist

    Hi Mike,

    Thanks for following up.

    Specific use case
    The idea is to let OAC’s AI Assistant enrich visualisations with live external context from MCP servers. For example:

    • In a CFO/ERP dashboard showing quarterly revenue by region, the Assistant could call an MCP server for foreign exchange rates to explain a margin drop (“UK revenue decline aligns with a 7% GBP/USD depreciation in Q3”).
    • In an HCM attrition dashboard, the Assistant could query an MCP server holding internal HR policy documents to highlight relevant changes (“Sales attrition linked to Feb 2025 compensation policy update”).

    This would save users from switching tools, and provide decision-ready narratives alongside the numbers.

    MCP servers I have in mind
    There are existing MCP servers that could deliver this context today:

    1. Alpha Vantage MCP Server – provides live and historical FX rates, stock movements, and market indicators for financial context.
    2. Financial Datasets MCP Server – delivers competitor benchmarks, industry index movements, and news sentiment for market comparison.
    3. Knowledge Base MCP Server (txtai-powered) – enables semantic search over internal policy, HR, or operational documents for procedural or compliance context.

    All three could be whitelisted in OAC for secure, auditable queries from the AI Assistant, without importing external data into the semantic model. There are many more but these are just to give you an idea of what I had in mind.