Support text to speech with WAV export

Add text-to-speech support so Objo desktop applications can read text aloud and save generated speech as a WAV audio file.

Applications should be able to choose an available voice, adjust speaking speed and volume, and stop speech. Speech synthesis and WAV export should run without freezing the application's interface and report completion or failure.

This would support read-aloud features, spoken prompts, and generating reusable audio from text.