docs: Add explicit VAST.ai support and clarify no-Docker deployment

docs: Update README for containerized environments like VST.ai

feat: Add complete microservices architecture for speech-to-speech avatar system