Provide Convenience Modules for Audio Processing

[PSchmiedmayer]

Problem

The current logic of mapping the raw Data output and input from the realtime models is not implemented in the main package but only shipped as part of the UI testing application: https://github.com/StanfordSpezi/SpeziLLM/tree/main/Tests/UITests/TestApp/LLMOpenAIRealtime

Solution

We should provide a convenient module that enables audio input and output modules as a nice wrapper similar to the implementation in the testing application with some additional structural improvements and a clear API surface.

Additional context

We might want to even remove the documentation requirements about the structure of the audio input and output. It is defined for the OpenAI Realtime API but might be different for other Realtime models.

Code of Conduct

• I agree to follow this project's Code of Conduct and Contributing Guidelines