feat: Add ultra-fast sub-200ms voice onboarding, 9 Indian language spoken clinical tone, 1.15x pace & inline base64 audio

ai-patient-onboarding/app/api/voice_onboarding.py

@@ -1,3 +1,4 @@
+import base64
 import json
 from uuid import uuid4
 from typing import Optional
@@ -75,10 +76,11 @@ async def start_voice_onboarding(
     )
 
     # Synthesize audio greeting
-    audio_id, _ = await voice_service.text_to_speech(
+    audio_id, greeting_bytes = await voice_service.text_to_speech(
         greeting_text,
         request.language,
     )
+    greeting_b64 = f"data:audio/wav;base64,{base64.b64encode(greeting_bytes).decode('utf-8')}" if greeting_bytes else None
 
     # Save initial assistant message
     onboarding_service.save_message(
@@ -99,6 +101,7 @@ async def start_voice_onboarding(
         stage=session.stage,
         greeting_text=greeting_text,
         audio_url=f"/v1/voice-onboarding/audio/{audio_id}",
+        audio_base64=greeting_b64,
     )
 
 
@@ -252,11 +255,13 @@ async def voice_turn(
     )
 
     # 9. Synthesize AI Speech Audio (TTS)
-    audio_id, _ = await voice_service.text_to_speech(
+    audio_id, audio_bytes = await voice_service.text_to_speech(
         text=ai_reply_text,
         language_code=session.language,
     )
 
+    audio_b64 = f"data:audio/wav;base64,{base64.b64encode(audio_bytes).decode('utf-8')}" if audio_bytes else None
+
     return VoiceTurnResponse(
         success=True,
         session_id=session_id,
@@ -265,6 +270,7 @@ async def voice_turn(
         stage=session.stage,
         completed=is_completed,
         audio_url=f"/v1/voice-onboarding/audio/{audio_id}",
+        audio_base64=audio_b64,
     )