Fix demos: build to /app/demo_samples, copy to DATA_DIR at runtime

Generate demo audio at Docker build time instead of runtime

Fix build: generate demos on first app startup instead of build time

Fix build error: restore fallback for demo generation

Debug: remove silent fallback from demo generation to see errors

Add Tier 1 improvements: Risk Score, KPIs, PDF, Demo Samples

Fix audio playback + persist logs across refreshes

Fix permission denied: store models/data outside /app mount

Fix permission denied on pretrained_models in HF Spaces

Fix HF Spaces deployment: use Docker SDK, pin huggingface_hub, pre-download models

Remove WebRTC, use upload only - more reliable

Fix WebRTC: add TURN servers for better connectivity

Force rebuild - fix st.dialog compatibility

Add SOP Audio Analyzer app files

Duplicate from streamlit/streamlit-template-space