Issue Fix

Upload 404 warning	doc_set(merge=True) in upload.py — creates doc if missing
MQTT call_end 404 error	doc_set(merge=True) in mqtt_handler.py — same root cause
Transcription 404 (saving transcript to nonexistent doc)	doc_set(merge=True) in transcription.py
Transcription ADC credentials error	Explicit service_account.Credentials from gcp-key.json in _sync_transcribe — same pattern as storage.py

drb-c2-core/app/internal/transcription.py

@@ -35,7 +35,7 @@ async def transcribe_call(call_id: str, gcs_uri: str) -> Optional[str]:
 
     if transcript:
         try:
-            await fstore.doc_update("calls", call_id, {"transcript": transcript})
+            await fstore.doc_set("calls", call_id, {"transcript": transcript})
             logger.info(f"Transcript saved for call {call_id} ({len(transcript)} chars)")
         except Exception as e:
             logger.warning(f"Could not save transcript for {call_id}: {e}")
@@ -46,8 +46,17 @@ async def transcribe_call(call_id: str, gcs_uri: str) -> Optional[str]:
 def _sync_transcribe(gcs_uri: str) -> Optional[str]:
     """Synchronous STT call — run in a thread via asyncio.to_thread."""
     from google.cloud import speech
+    from app.config import settings
 
-    client = speech.SpeechClient()
+    if settings.gcp_credentials_path:
+        from google.oauth2 import service_account
+        creds = service_account.Credentials.from_service_account_file(
+            settings.gcp_credentials_path,
+            scopes=["https://www.googleapis.com/auth/cloud-platform"],
+        )
+        client = speech.SpeechClient(credentials=creds)
+    else:
+        client = speech.SpeechClient()
 
     audio = speech.RecognitionAudio(uri=gcs_uri)
     config = speech.RecognitionConfig(