Process Audio for Chatbots

This is a set of 6 Semantic Kernel Python concept samples demonstrating audio processing patterns: chat with audio input (01-chat_with_audio_input.py), chat with audio output (02-chat_with_audio_output.py), chat with both audio input and output (03-chat_with_audio_input_output.py), generating audio directly from a prompt (audio_from_prompt.py), and utility scripts for playing (audio_player.py) and recording (audio_recorder.py) audio.

Use these samples as reference implementations when building a Semantic Kernel chat agent that needs to accept spoken input, respond with generated speech, or both - covering the progression from text-only chat to full voice-in/voice-out interaction. The prompt-to-audio and player/recorder utilities are useful building blocks even outside a full chat context.

Not a single unified audio pipeline - these are six separate, focused concept examples meant to be studied and adapted individually rather than run as one integrated application out of the box.

Inputs: audio input (microphone/recorded audio) for the input-handling samples, or text prompts for the audio-generation sample.

Outputs: transcribed text (speech-to-text samples), generated speech audio (text-to-speech samples), or played/recorded audio files (utility scripts).

Microsoft Semantic Kernel (Python), with speech-to-text and text-to-speech model backends handled via Semantic Kernel's audio abstractions.

Developers building voice-enabled Semantic Kernel agents who need working reference code for audio input, audio output, or both, plus basic audio playback/recording utilities.