Move from data to intelligence with Oracle MCP and Microsoft IQ | ODSP919
Jeff Smith and Ram Kakani present an architecture and demo that connects Oracle data running in Oracle Database@Azure to Microsoft’s AI “intelligence layer” using Oracle managed MCP Servers and Microsoft IQ.
Overview
The session focuses on moving from disconnected enterprise data to AI-driven, agentic workflows by combining:
- Oracle managed MCP Servers to provide connectivity to Oracle data sources (including Oracle Database@Azure)
- Microsoft IQ (Work IQ, Fabric IQ, Foundry IQ) to add context, reasoning, orchestration, and governance
Architecture: Oracle MCP + Microsoft IQ
Oracle MCP Servers for data connectivity
- MCP is presented as the connectivity mechanism that lets an AI system access enterprise data and tools through managed servers.
- Oracle’s managed MCP Servers are positioned as handling infrastructure and operations for the MCP layer.
- The goal is to accelerate time-to-value while keeping the solution enterprise-ready and scalable.
Microsoft IQ as the intelligence layer
Microsoft IQ is described as the layer that brings:
- Context over enterprise data
- Reasoning to interpret and act on that context
- Orchestration to coordinate multi-step workflows
- Governance features (called out alongside “Agent ID” and “Agent 365” in the session outline)
Development stack components highlighted
The session outline calls out a development stack that includes:
- Azure AI Foundry capabilities
- Foundry agent services (agent service layer)
- GitHub Copilot as part of the developer toolchain
Setup flow: MCP server connectivity and authentication
The presenters walk through a step-by-step setup sequence that includes:
- Establishing MCP server connectivity
- Configuring authentication for access to Oracle data through the MCP layer
Demo: Oracle Database@Azure connected through MCP
Environment setup
- A live setup sequence is shown using the Azure portal.
- The demo centers on integrating Azure-side components with Oracle databases via Oracle Database@Azure.
Running an agent over enterprise data
- The agent is run to analyze a business scenario: unpaid invoices.
- The workflow uses Foundry IQ together with Oracle MCP connectivity to retrieve and reason over the data.
Recommendations and workflow actions
- The agent produces actionable recommendations based on the invoice analysis.
- The session outline describes follow-on workflow automation using Work IQ.
Closing and resources
- The session closes with community resources and an invitation to explore Oracle at Azure integrations.
- It is identified as a Microsoft Build 2026 session (ODSP919).