feat: implement AI-orchestrated result extraction and service connectivity tests

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

@@ -130,9 +130,14 @@ async def voice_turn(
             detail="Voice onboarding session is not active",
         )
 
-    # 2. Transcribe speech if audio provided with multi-model Indic recognition
+    # 2. Fast-path transcript: use real-time Web Speech text_fallback first for zero-latency processing
     patient_transcript = ""
-    if audio:
+    if text_fallback and text_fallback.strip():
+        from app.services.medical_speech_normalizer import normalize_medical_speech
+        patient_transcript = normalize_medical_speech(text_fallback.strip(), language_code=session.language)
+
+    # If text_fallback was not available, run server-side multi-model STT
+    if not patient_transcript and audio:
         audio_bytes = await audio.read()
         if len(audio_bytes) > 0:
             patient_transcript = await voice_service.speech_to_text(
@@ -147,18 +152,35 @@ async def voice_turn(
                 },
             )
 
-    # If backend STT is empty, use frontend real-time transcript fallback
-    if (not patient_transcript or not patient_transcript.strip()) and text_fallback and text_fallback.strip():
-        from app.services.medical_speech_normalizer import normalize_medical_speech
-        patient_transcript = normalize_medical_speech(text_fallback.strip(), language_code=session.language)
-
+    # If STILL empty, user spoke nothing or silence was recorded.
+    # Do NOT substitute fake default text or auto-fill clinical fields!
     if not patient_transcript or not patient_transcript.strip():
-        if session.language.startswith("ta"):
-            patient_transcript = "எனக்கு உடல்நலக்குறைவு உள்ளது."
-        elif session.language.startswith("hi"):
-            patient_transcript = "मुझे स्वास्थ्य संबंधी समस्या है।"
+        lang = session.language.lower()
+        if lang.startswith("ta"):
+            no_speech_msg = "மன்னிக்கவும், உங்கள் குரல் கேட்கவில்லை. தயவுசெய்து உங்கள் உடல்நலப் பிரச்சினை பற்றி பேசவும்."
+        elif lang.startswith("hi"):
+            no_speech_msg = "क्षमा करें, मुझे आपकी आवाज़ सुनाई नहीं दी। कृपया अपनी स्वास्थ्य समस्या के बारे में बोलें।"
+        elif lang.startswith("te"):
+            no_speech_msg = "క్షమించండి, మీ ధ్వని వినిపించలేదు. దయచేసి మీ ఆరోగ్య సమస్య గురించి మాట్లాడండి."
+        elif lang.startswith("kn"):
+            no_speech_msg = "ಕ್ಷಮಿಸಿ, ನಿಮ್ಮ ಧ್ವನಿ ಕೇಳಿಸಲಿಲ್ಲ. ದಯವಿಟ್ಟು ನಿಮ್ಮ ಆರೋಗ್ಯದ ತೊಂದರೆ ಬಗ್ಗೆ ಮಾತನಾಡಿ."
         else:
-            patient_transcript = "I am experiencing symptoms."
+            no_speech_msg = "I didn't hear anything. Please speak into your microphone and describe your health symptoms."
+
+        audio_id, _ = await voice_service.text_to_speech(
+            text=no_speech_msg,
+            language_code=session.language,
+        )
+
+        return VoiceTurnResponse(
+            success=True,
+            session_id=session_id,
+            patient_transcript="",
+            ai_text=no_speech_msg,
+            stage=session.stage,
+            completed=False,
+            audio_url=f"/v1/voice-onboarding/audio/{audio_id}",
+        )