🔧(summary) add configurable language settings for WhisperX transcription

Make WhisperX language detection configurable through FastAPI settings
to handle empty audio start scenarios where automatic detection fails and
incorrectly defaults to English despite 99% French usage.

Quick fix acknowledging long-term solution should allow dynamic
per-recording language selection configured by users through web
interface rather than global server settings.

env.d/development/summary.dist

@@ -11,6 +11,7 @@ AWS_S3_SECRET_ACCESS_KEY="password"
 WHISPERX_BASE_URL="https://configure-your-url.com"
 WHISPERX_ASR_MODEL="large-v2"
 WHISPERX_API_KEY="your-secret-key"
+WHISPERX_DEFAULT_LANGUAGE="fr"
 
 LLM_BASE_URL="https://configure-your-url.com"
 LLM_API_KEY="dev-apikey"

src/helm/env.d/dev-keycloak/values.meet.yaml.gotmpl

@@ -155,6 +155,7 @@ summary:
     WHISPERX_API_KEY: your-secret-value
     WHISPERX_BASE_URL: https://configure-your-url.com
     WHISPERX_ASR_MODEL: large-v2
+    WHISPERX_DEFAULT_LANGUAGE: fr
     LLM_BASE_URL: https://configure-your-url.com
     LLM_API_KEY: your-secret-value
     LLM_MODEL: meta-llama/Llama-3.1-8B-Instruct
@@ -191,6 +192,7 @@ celeryTranscribe:
     WHISPERX_API_KEY: your-secret-value
     WHISPERX_BASE_URL: https://configure-your-url.com
     WHISPERX_ASR_MODEL: large-v2
+    WHISPERX_DEFAULT_LANGUAGE: fr
     LLM_BASE_URL: https://configure-your-url.com
     LLM_API_KEY: your-secret-value
     LLM_MODEL: meta-llama/Llama-3.1-8B-Instruct
@@ -228,6 +230,7 @@ celerySummarize:
     WHISPERX_API_KEY: your-secret-value
     WHISPERX_BASE_URL: https://configure-your-url.com
     WHISPERX_ASR_MODEL: large-v2
+    WHISPERX_DEFAULT_LANGUAGE: fr
     LLM_BASE_URL: https://configure-your-url.com
     LLM_API_KEY: your-secret-value
     LLM_MODEL: meta-llama/Llama-3.1-8B-Instruct

src/summary/summary/core/celery_worker.py

@@ -270,7 +270,9 @@ def process_audio_transcribe_summarize_v2(
         transcription_start_time = time.time()
         with open(temp_file_path, "rb") as audio_file:
             transcription = whisperx_client.audio.transcriptions.create(
-                model=settings.whisperx_asr_model, file=audio_file
+                model=settings.whisperx_asr_model,
+                file=audio_file,
+                language=settings.whisperx_default_language,
             )
             metadata_manager.track(
                 task_id,

src/summary/summary/core/config.py

@@ -39,6 +39,8 @@ class Settings(BaseSettings):
     whisperx_base_url: str = "https://api.openai.com/v1"
     whisperx_asr_model: str = "whisper-1"
     whisperx_max_retries: int = 0
+    # ISO 639-1 language code (e.g., "en", "fr", "es")
+    whisperx_default_language: Optional[str] = None
     llm_base_url: str
     llm_api_key: str
     llm_model: str