Update app.py

app.py

@@ -95,22 +95,59 @@ class MockInterviewer:
     def _create_cache_key(self, job_role: str, company: str) -> str:
         return f'{job_role.lower()}+{company.lower()}'
 
+    def transcribe_and_chat(self, audio_file, job_role: str, company: str):
+        # Transcribe audio
+        transcript = self.transcribe_audio(audio_file)
+        # Now proceed with chat using the transcript
+        # Assuming chat method can handle initial user message as text
+        response = next(self.chat({'text': transcript}, [], job_role, company))
+        return response
+
+    def transcribe_audio(self, audio_file_path):
+        # Read the audio file
+        with open(audio_file_path, "rb") as audio_file:
+            audio_data = audio_file.read()
+
+        # Use OpenAI's API to transcribe the audio
+        response = openai.Speech.create(
+            audio=audio_data,
+            model="whisper-large"  # or whichever model is preferred
+        )
+
+        # Extract and return the transcription text
+        transcript = response['data']['text']
+        return transcript
+
 # Creating the Gradio interface
 with gr.Blocks() as demo:
     mock_interviewer = MockInterviewer()
 
+    with gr.Row():
+        job_role = gr.Textbox(label='Job Role', placeholder='Product Manager')
+        company = gr.Textbox(label='Company', placeholder='Amazon')
+        usr_audio = gr.Audio(source="microphone", type="filepath", label="Record or Upload Audio")
+
+    submit_btn = gr.Button("Submit")
+
+    response_output = gr.Textbox(label="Interviewer Response")
+
+    # Use transcribe_and_chat when audio is provided
+    submit_btn.click(
+        fn=mock_interviewer.transcribe_and_chat,
+        inputs=[usr_audio, job_role, company],
+        outputs=response_output
+    )
+
     chat_interface = gr.ChatInterface(
-        mock_interviewer.chat,
-        additional_inputs=[
-            gr.Textbox(label='Job Role', placeholder='Product Manager'),
-            gr.Textbox(label='Company', placeholder='Amazon')
-        ], 
-        title='I am your AI mock interviewer', 
+        fn=lambda usr_message, history, job_role, company: mock_interviewer.chat(usr_message, history, job_role, company),
+        additional_inputs=[job_role, company],
+        title='I am your AI mock interviewer',
         description='Make your selections above to configure me.',
         multimodal=True,
         retry_btn=None,
-        undo_btn=None).queue()
-    
+        undo_btn=None
+    ).queue()
+
     chat_interface.load(mock_interviewer.clear_thread)
     chat_interface.clear_btn.click(mock_interviewer.clear_thread)