Build enterprise .NET agents with Oracle AI Database RAG and Microsoft’s newest AI framework

Key Takeaways

  • Oracle AI Database Vector Store Connector now supports Microsoft Agent Framework, letting .NET developers use the database as a vector store.
  • The connector allows .NET apps to use semantic search with Oracle AI Database retrieval-augmented generation and create, read, update, and delete vectors in Oracle AI Database.
  • Developers can use it for enterprise-grade agents, LLM apps, MCP solutions, workflows, and retrieval-augmented generation.
Diagram illustrating a .NET application connecting bidirectionally between Oracle Database and Microsoft Semantic Kernel, showing the application acting as the integration layer between the database and AI framework.
.NET Application Integration Architecture

The Oracle AI Database Vector Store Connector (Oracle.VectorData) now supports Microsoft Agent Framework 1.7 and later, so .NET developers can use Oracle AI Database 26ai as a vector store for AI-based applications.

The connector lets .NET apps search, create, read, update, and delete vectors in Oracle AI Database. Use it to build agentic, language model (LLM), Model Context Protocol (MCP), and workflow solutions that rely on native AI features in ODP.NET and Oracle AI Database. Download the Oracle connector from NuGet Gallery.

The connector brings the Oracle data, vectors, and vector indexes to Microsoft Agent Framework services. In an earlier post about the Oracle AI Database Vector Store Connector, I covered its main features, .NET vector and Semantic Kernel support, and a comprehensive sample application demonstrating its capabilities. Those features also work with Microsoft Agent Framework.

Using Oracle AI Database Vector Store Connector with Microsoft Agent Framework

Microsoft Agent Framework is an SDK and runtime for building, orchestrating, and deploying AI agents and multi-agent workflows. It combines AutoGen’s simple agent patterns with Semantic Kernel’s enterprise features and was first released in production in April.

Microsoft Agent Framework, like any AI technology, requires data to operate, learn, and adapt. It uses Microsoft’s published connector model to manage and access data in a vector store, which is typically a database. In our case, the Oracle AI Database Vector Store Connector links Microsoft Agent Framework and Oracle AI Database vectors and data for retrieval-augmented generation (RAG).

Because the connector model is designed to work with multiple data sources with minimal to no data source specific code required, it’s easy to incorporate existing Microsoft Agent Framework, .NET vector, and Semantic Kernel apps with Oracle AI Database 26ai into existing AI apps.

Oracle AI Database AI Capabilities

Oracle AI Database 26ai is built to solve enterprise AI challenges. With Oracle AI Vector Search, .NET developers can add AI-powered similarity search to structured and unstructured data while keeping the database’s security, consistency, and operational features. Applications can combine semantic search with exact-match queries to support more precise AI workflows.

Oracle AI Vector Search can work with many LLMs. With RAG, .NET apps can combine an LLM with your database’s business data to return more relevant, better-grounded results.

In this video, I present the Oracle vector store capabilities, then demo an AI chat app that uses RAG from an Oracle AI Database with Microsoft Agent Framework. The sample app uses two agents for its RAG: one for Oracle AI Database and one for Microsoft Foundry.

Building .NET AI Apps with Oracle AI Database

Whether you’re building one assistant or coordinating specialized agents in a workflow, Oracle AI Database and Microsoft Agent Framework provide .NET apps multi-agent orchestration, multi-provider model support, and MCP interoperability. Start developing AI apps with these technologies today!

Learn more

FAQs

  • What does the Oracle AI Database Vector Store Connector do?
    It connects applications to Oracle AI Database so they can manage and search vector data using Microsoft Agent Framework, .NET vectors, and Semantic Kernel.
  • Who is this update for?
    This update is for .NET developers building AI assistants, agent workflows, LLM apps, or MCP-based solutions with the new Microsoft Agent Framework.
  • When will the Oracle AI Database Vector Store Connector be production?
    Oracle plans to release a production .NET vector store connector around the same time as when Microsoft releases its .NET vector store connectors in production. Currently, they are all in preview.
  • What software components do I need to create an app with the Oracle connector and Microsoft Agent Framework?
    Download Oracle.VectorData from NuGet Gallery, add Microsoft Agent Framework 1.7 or higher to your project, and connect to Oracle AI Database 26ai. The database can be on-premises, cloud-based, or autonomous.