Integrate Multimodal Language Models into .NET Backends with OpenAI .NET API and .NET Aspire

This talk was given at the October 2024 Generation AI Meetup in Karlsruhe, Germany.

Prerequisites

You require the following prerequisites:

• .NET 8 SDK
• An OpenAI API key

You can configure the API key by creating a file called appsettings.Development.json in the folder backend/AiInformationExtractionApi and add the following contents:

{
    "openAi": {
        "apiKey": "paste your OpenAI API key here"
    }
}

You can also use dotnet user-secrets to achieve the same thing.

How to compile and run the example

Simply build and run the solution via your IDE or with dotnet run.

The most important integration tests can be found in the AspireAppHost.IntegrationTests project:

• TextIntegrationTests
• SpeechIntegrationTests
• ImageIntegrationTests