adds ability to pass in api key for every transcriber agent and synthesizer

vocode/streaming/synthesizer/azure_synthesizer.py

@@ -52,14 +52,27 @@ class AzureSynthesizer(BaseSynthesizer):
     OFFSET_MS = 100
 
     def __init__(
-        self, synthesizer_config: AzureSynthesizerConfig, logger: logging.Logger = None
+        self,
+        synthesizer_config: AzureSynthesizerConfig,
+        logger: logging.Logger = None,
+        azure_speech_key: str = None,
+        azure_speech_region: str = None,
     ):
         super().__init__(synthesizer_config)
         self.synthesizer_config = synthesizer_config
         # Instantiates a client
+        azure_speech_key = azure_speech_key or getenv("AZURE_SPEECH_KEY")
+        azure_speech_region = azure_speech_region or getenv("AZURE_SPEECH_REGION")
+        if not azure_speech_key:
+            raise ValueError(
+                "Please set AZURE_SPEECH_KEY environment variable or pass it as a parameter"
+            )
+        if not azure_speech_region:
+            raise ValueError(
+                "Please set AZURE_SPEECH_REGION environment variable or pass it as a parameter"
+            )
         speech_config = speechsdk.SpeechConfig(
-            subscription=getenv("AZURE_SPEECH_KEY"),
-            region=getenv("AZURE_SPEECH_REGION"),
+            subscription=azure_speech_key, region=azure_speech_region
         )
         if self.synthesizer_config.audio_encoding == AudioEncoding.LINEAR16:
             if self.synthesizer_config.sampling_rate == 44100:

vocode/streaming/synthesizer/eleven_labs_synthesizer.py

@@ -24,7 +24,7 @@ ADAM_VOICE_ID = "pNInz6obpgDQGcFmaJgB"
 class ElevenLabsSynthesizer(BaseSynthesizer):
     def __init__(self, config: ElevenLabsSynthesizerConfig):
         super().__init__(config)
-        self.api_key = getenv("ELEVEN_LABS_API_KEY")
+        self.api_key = config.api_key or getenv("ELEVEN_LABS_API_KEY")
         self.voice_id = config.voice_id or ADAM_VOICE_ID
         self.words_per_minute = 150
 

vocode/streaming/synthesizer/google_synthesizer.py

@@ -3,6 +3,7 @@ import wave
 from typing import Any, Optional
 
 from google.cloud import texttospeech_v1beta1 as tts
+from vocode import getenv
 
 from vocode.streaming.agent.bot_sentiment_analyser import BotSentiment
 from vocode.streaming.models.message import BaseMessage
@@ -22,6 +23,10 @@ class GoogleSynthesizer(BaseSynthesizer):
     def __init__(self, synthesizer_config: GoogleSynthesizerConfig):
         super().__init__(synthesizer_config)
         # Instantiates a client
+        if not getenv("GOOGLE_APPLICATION_CREDENTIALS"):
+            raise Exception(
+                "GOOGLE_APPLICATION_CREDENTIALS environment variable must be set"
+            )
         self.client = tts.TextToSpeechClient()
 
         # Build the voice request, select the language code ("en-US") and the ssml