update

DOCKERFILE

@@ -4,20 +4,24 @@ FROM python:3.11-slim
 # Set the working directory in the container
 WORKDIR /app
 
-# Copy the requirements file into the container
-COPY requirements.txt .
-
-# Install any needed system dependencies (like for audio)
+# Install system dependencies
 RUN apt-get update && apt-get install -y --no-install-recommends \
     ffmpeg \
+    git \
     && rm -rf /var/lib/apt/lists/*
 
+# Copy the requirements file into the container
+COPY requirements.txt .
+
 # Install the Python dependencies
 RUN pip install --no-cache-dir -r requirements.txt
 
 # Copy the rest of the application's code into the container
 COPY . .
 
+# Create necessary directories
+RUN mkdir -p uploads reports
+
 # Expose the port that Gradio runs on
 EXPOSE 7860
 

README_HF.md

@@ -0,0 +1,28 @@
+# AI Interview Coach - Hugging Face Spaces Deployment
+
+This is an AI-powered interview coaching application that helps users practice for Product Manager interviews.
+
+## Features
+- Interactive voice-based interview simulation
+- Document upload for personalized questions
+- Real-time feedback and evaluation
+- PDF report generation
+
+## Setup for Hugging Face Spaces
+
+### Required Environment Variables
+Add these secrets in your Hugging Face Space settings:
+
+```
+GROQ_API_KEY=your_groq_api_key_here
+```
+
+### Known Limitations in HF Spaces
+- TTS (Text-to-Speech) audio generation may be disabled due to system dependencies
+- Some audio features might not work in the containerized environment
+
+## Local Development
+To run locally, ensure you have:
+1. A valid GROQ API key set as environment variable
+2. All required system dependencies installed
+3. The voice model files in the `voice_model/` directory

app.py

@@ -36,11 +36,18 @@ def start_interview(interview_type, doc_file, name, num_questions):
     initial_state["current_question_text"] = first_question
     greeting = f"Hello {initial_state['name']}. We'll go through {int(num_questions)} questions today. Here is your first question:"
     tts_prompt = f"{greeting} {first_question}"
-    ai_voice_path = text_to_speech_file(tts_prompt)
+    
+    # Try to generate TTS audio, but don't fail if it's not available
+    try:
+        ai_voice_path = text_to_speech_file(tts_prompt)
+    except Exception as e:
+        print(f"TTS generation failed: {e}")
+        ai_voice_path = None
+    
     return {
         state: initial_state,
         chatbot: gr.update(value=[[None, f"{greeting}\n\n{first_question}"]]),
-        audio_out: gr.update(value=ai_voice_path, autoplay=True),
+        audio_out: gr.update(value=ai_voice_path, autoplay=True if ai_voice_path else False),
         audio_in: gr.update(interactive=True),
         start_btn: gr.update(interactive=False)
     }
@@ -60,10 +67,17 @@ def handle_interview_turn(user_audio, chatbot_history, current_state):
         end_message = "This concludes the interview. Generating your final report now."
         chatbot_history.append([None, end_message])
         pdf_path = generate_pdf_file(current_state)
-        ai_voice_path = text_to_speech_file(end_message)
+        
+        # Try to generate TTS audio, but don't fail if it's not available
+        try:
+            ai_voice_path = text_to_speech_file(end_message)
+        except Exception as e:
+            print(f"TTS generation failed: {e}")
+            ai_voice_path = None
+            
         yield {
             chatbot: chatbot_history,
-            audio_out: gr.update(value=ai_voice_path, autoplay=True),
+            audio_out: gr.update(value=ai_voice_path, autoplay=True if ai_voice_path else False),
             download_pdf_btn: gr.update(value=pdf_path, visible=True)
         }
     else:
@@ -73,11 +87,18 @@ def handle_interview_turn(user_audio, chatbot_history, current_state):
         q_num = current_state["current_question_num"]
         transition_message = f"Thank you. Here is question {q_num}:\n\n{next_question}"
         chatbot_history.append([None, transition_message])
-        ai_voice_path = text_to_speech_file(transition_message)
+        
+        # Try to generate TTS audio, but don't fail if it's not available
+        try:
+            ai_voice_path = text_to_speech_file(transition_message)
+        except Exception as e:
+            print(f"TTS generation failed: {e}")
+            ai_voice_path = None
+            
         yield {
             state: current_state,
             chatbot: chatbot_history,
-            audio_out: gr.update(value=ai_voice_path, autoplay=True),
+            audio_out: gr.update(value=ai_voice_path, autoplay=True if ai_voice_path else False),
             audio_in: gr.update(interactive=True)
         }
 
@@ -133,4 +154,9 @@ with gr.Blocks(theme=gr.themes.Default()) as app:
 if __name__ == "__main__":
     os.makedirs(config.UPLOAD_FOLDER, exist_ok=True)
     os.makedirs(config.REPORT_FOLDER, exist_ok=True)
-    app.launch(debug=True)
+    # Configure for Hugging Face Spaces deployment
+    app.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        share=False
+    )

check_env.py

@@ -0,0 +1,61 @@
+#!/usr/bin/env python3
+"""
+Environment validation script for AI Interview Coach
+Run this before deploying to check for common issues
+"""
+
+import os
+import sys
+
+def check_environment():
+    """Check if the environment is properly configured"""
+    issues = []
+    
+    # Check for required environment variables
+    if not os.environ.get("GROQ_API_KEY"):
+        issues.append("❌ GROQ_API_KEY environment variable is not set")
+    else:
+        print("✅ GROQ_API_KEY is set")
+    
+    # Check for required directories
+    required_dirs = ['uploads', 'reports']
+    for dir_name in required_dirs:
+        if not os.path.exists(dir_name):
+            issues.append(f"❌ Directory '{dir_name}' does not exist")
+        else:
+            print(f"✅ Directory '{dir_name}' exists")
+    
+    # Check for voice model file
+    voice_model_path = './voice_model/en_US-lessac-medium.onnx'
+    if not os.path.exists(voice_model_path):
+        issues.append(f"⚠️  Voice model file not found at {voice_model_path} (TTS will be disabled)")
+    else:
+        print("✅ Voice model file found")
+    
+    # Try to import critical modules
+    try:
+        import gradio
+        print("✅ Gradio imported successfully")
+    except ImportError:
+        issues.append("❌ Gradio not installed")
+    
+    try:
+        from groq import Groq
+        print("✅ Groq imported successfully")
+    except ImportError:
+        issues.append("❌ Groq not installed")
+    
+    # Summary
+    if issues:
+        print("\n🚨 Issues found:")
+        for issue in issues:
+            print(f"  {issue}")
+        print(f"\nFound {len(issues)} issue(s) that need to be addressed.")
+        return False
+    else:
+        print("\n🎉 All checks passed! Ready for deployment.")
+        return True
+
+if __name__ == "__main__":
+    success = check_environment()
+    sys.exit(0 if success else 1)

config.py

@@ -8,7 +8,7 @@ OLLAMA_MODEL = 'llama3.1'
 INTERVIEW_TYPES = ['Product Sense', 'Technical', 'General Product Interview', 'Group Discussion (GD)', 'Root case analysis']
 
 # -- Piper TTS Configuration --
-PIPER_VOICE_MODEL = './voice_model/en_US-lessac-medium.onnx' 
+PIPER_VOICE_MODEL = os.path.join(os.path.dirname(__file__), 'voice_model', 'en_US-lessac-medium.onnx')
 
 # -- Directories --
 UPLOAD_FOLDER = 'uploads'

modules/llm_handler.py

@@ -6,7 +6,11 @@ from groq import Groq
 from modules.web_search import search_for_example_answers
 
 # Initialize the Groq client
-client = Groq(api_key=os.environ.get("GROQ_API_KEY"))
+groq_api_key = os.environ.get("GROQ_API_KEY")
+if not groq_api_key:
+    raise ValueError("GROQ_API_KEY environment variable is required but not set")
+
+client = Groq(api_key=groq_api_key)
 MODEL = "llama3-70b-8192" # Use the more powerful 70B model for detailed analysis
 
 def generate_question(interview_type, document_text):

modules/stt_handler.py

@@ -12,18 +12,22 @@ def transcribe_audio(audio_filepath):
         with sr.AudioFile(audio_filepath) as source:
             audio_data = recognizer.record(source)
             print("Transcribing with Whisper...")
+            # Use Whisper for transcription
             text = recognizer.recognize_whisper(audio_data, language="english")
             print(f"User transcribed as: {text}")
             return text
     except sr.UnknownValueError:
         print("STT Error: Whisper could not understand the audio.")
-        return "[Could not understand audio]"
+        return "[Could not understand audio - please try speaking more clearly]"
     except sr.RequestError as e:
         print(f"STT Error: Could not request results from Whisper service; {e}")
-        return f"[Transcription error: Whisper service issue - {e}]"
+        return "[Transcription service temporarily unavailable - please try again]"
+    except ImportError as e:
+        print(f"STT Error: Whisper not available: {e}")
+        return "[Speech recognition not available in this environment]"
     except Exception as e:
         print(f"STT Error: An unexpected error occurred during transcription: {e}")
-        return f"[Transcription error: {e}]"
+        return f"[Transcription error: Please try again]"
     finally:
         if os.path.exists(audio_filepath):
             try:

modules/tts_handler.py

@@ -9,8 +9,14 @@ import tempfile
 
 def text_to_speech_file(text_to_speak):
     print(f"AI generating audio for: {text_to_speak}")
-    piper_executable = 'piper'  # Use system PATH
+    
+    # For Hugging Face Spaces deployment, we'll disable TTS audio generation
+    # since piper-tts requires system dependencies that may not be available
     try:
+        # Check if piper executable exists
+        piper_executable = 'piper'
+        
+        # Try to use piper if available, otherwise skip audio generation
         with tempfile.NamedTemporaryFile(delete=False, suffix=".raw") as raw_file:
             raw_filename = raw_file.name
 
@@ -27,5 +33,6 @@ def text_to_speech_file(text_to_speak):
         os.remove(raw_filename)
         return wav_filename
     except Exception as e:
-        print(f"An error occurred during TTS generation: {e}")
+        print(f"TTS not available in this environment: {e}")
+        # Return None to disable audio playback in deployment
         return None

requirements.txt

@@ -1,16 +1,16 @@
-ollama
+gradio==4.44.0
 openai-whisper
-gradio
 pydub
 soundfile
-pyaudio
-piper-tts
 PyMuPDF
 python-docx
 reportlab
 speechrecognition
 duckduckgo-search
-ddgs
 matplotlib
 regex
-groq
+groq
+# Removed problematic dependencies for HF Spaces:
+# - ollama (local service, not available in HF Spaces)
+# - pyaudio (often causes build issues)
+# - piper-tts (system dependencies issues)