Update app.py

app.py

@@ -5,10 +5,10 @@ import io
 import os
 import tempfile
 from PIL import Image, ImageDraw, ImageFont
+import requests
 import json
 from datetime import datetime
-import re
-import random
+import time
 
 # Import with error handling
 try:
@@ -16,7 +16,6 @@ try:
     TRANSFORMERS_AVAILABLE = True
 except ImportError:
     TRANSFORMERS_AVAILABLE = False
-    st.error("Install transformers: pip install transformers")
 
 try:
     import google.generativeai as genai
@@ -34,437 +33,688 @@ except ImportError:
 st.set_page_config(
     page_title="VoiceCanvas - AI Content Studio",
     page_icon="🎨",
-    layout="wide"
+    layout="wide",
+    initial_sidebar_state="expanded"
 )
 
 # Initialize session state
-def init_session_state():
-    defaults = {
-        'generated_content': {},
-        'transcription': "",
-        'models_loaded': False,
-        'whisper_model': None,
-        'processing': False
-    }
-    for key, value in defaults.items():
-        if key not in st.session_state:
-            st.session_state[key] = value
-
-init_session_state()
+if 'generated_content' not in st.session_state:
+    st.session_state.generated_content = {}
+if 'transcription' not in st.session_state:
+    st.session_state.transcription = ""
+if 'processing' not in st.session_state:
+    st.session_state.processing = False
+if 'current_task' not in st.session_state:
+    st.session_state.current_task = ""
+if 'models_loaded' not in st.session_state:
+    st.session_state.models_loaded = False
+if 'whisper_model' not in st.session_state:
+    st.session_state.whisper_model = None
+if 'button_clicked' not in st.session_state:
+    st.session_state.button_clicked = False
 
-@st.cache_resource
-def load_whisper_model():
-    """Load Whisper model with caching"""
+def load_models():
+    """Load models efficiently with progress tracking"""
+    
+    if st.session_state.models_loaded and st.session_state.whisper_model is not None:
+        return True
+    
     if not TRANSFORMERS_AVAILABLE:
-        return None
+        st.error("❌ Transformers library not available. Please install: pip install transformers")
+        return False
+    
+    progress_bar = st.progress(0)
+    status_text = st.empty()
     
     try:
-        model = pipeline(
+        # Load Whisper model
+        status_text.text("Loading speech recognition model...")
+        progress_bar.progress(25)
+        
+        # Use session state to store the model
+        st.session_state.whisper_model = pipeline(
             "automatic-speech-recognition",
             model="openai/whisper-tiny",
-            device=-1,
+            device=-1,  # Use CPU
             torch_dtype=torch.float32,
             return_timestamps=False
         )
-        return model
+        
+        progress_bar.progress(75)
+        status_text.text("Models loaded successfully!")
+        progress_bar.progress(100)
+        
+        st.session_state.models_loaded = True
+        
+        # Clear progress indicators after a moment
+        time.sleep(1)
+        progress_bar.empty()
+        status_text.empty()
+        
+        return True
+        
     except Exception as e:
-        st.error(f"Model loading error: {e}")
-        return None
+        st.error(f"❌ Error loading models: {str(e)}")
+        st.error("Try installing additional dependencies: pip install librosa soundfile")
+        progress_bar.empty()
+        status_text.empty()
+        return False
 
 def setup_gemini():
-    """Setup Gemini API"""
+    """Setup Gemini API if available"""
     if not GENAI_AVAILABLE:
         return False
-    
-    api_key = os.getenv("GEMINI_API_KEY") or st.secrets.get("GEMINI_API_KEY", "") if hasattr(st, 'secrets') else ""
-    
-    if api_key:
-        genai.configure(api_key=api_key)
-        return True
-    return False
-
-def transcribe_audio(audio_input):
-    """Transcribe audio to text"""
+        
     try:
-        model = load_whisper_model()
-        if not model:
-            return "Error: Speech recognition not available"
+        api_key = os.getenv("GEMINI_API_KEY")
+        if not api_key and hasattr(st, 'secrets'):
+            api_key = st.secrets.get("GEMINI_API_KEY", "")
         
-        result = model(audio_input)
-        return result.get("text", "").strip() if isinstance(result, dict) else str(result).strip()
-    
+        if api_key:
+            genai.configure(api_key=api_key)
+            return True
+        return False
     except Exception as e:
-        return f"Transcription error: {str(e)}"
-
-def extract_keywords(text):
-    """Extract key terms from user input"""
-    # Remove common words
-    stop_words = {'the', 'a', 'an', 'and', 'or', 'but', 'in', 'on', 'at', 'to', 'for', 'of', 'with', 'by', 'is', 'are', 'was', 'were', 'will', 'would', 'could', 'should', 'can', 'may', 'might', 'this', 'that', 'these', 'those'}
-    words = re.findall(r'\b[a-zA-Z]{3,}\b', text.lower())
-    keywords = [word for word in words if word not in stop_words]
-    return list(set(keywords))[:10]  # Top 10 unique keywords
+        return False
 
-def generate_dynamic_content(user_input):
-    """Generate dynamic content based on user input"""
-    keywords = extract_keywords(user_input)
-    input_words = user_input.split()
-    
-    # Extract potential product/service name (first significant noun)
-    product_name = next((word for word in input_words if len(word) > 4 and word.lower() not in ['that', 'with', 'this', 'your', 'their', 'could', 'would', 'should']), "your product")
-    
-    # Dynamic taglines based on user input
-    tagline_templates = [
-        f"Revolutionize your {keywords[0] if keywords else 'experience'} with {product_name}",
-        f"{product_name.title()}: Where {keywords[1] if len(keywords) > 1 else 'innovation'} meets {keywords[2] if len(keywords) > 2 else 'excellence'}",
-        f"Transform {keywords[0] if keywords else 'your world'} - Choose {product_name.title()}"
-    ]
-    
-    # Dynamic social media posts
-    hashtags = ['#' + word.title() for word in keywords[:3]]
-    social_posts = [
-        f"🚀 Ready to experience {product_name}? Discover how it's changing {keywords[0] if keywords else 'everything'}! Join thousands already benefiting. {' '.join(hashtags[:2])}",
-        f"💡 What makes {product_name} special? It's all about {keywords[1] if len(keywords) > 1 else 'innovation'} and {keywords[2] if len(keywords) > 2 else 'quality'}. See the difference today! {' '.join(hashtags[1:3])}",
-        f"✨ {product_name.title()} users report amazing results with {keywords[0] if keywords else 'their experience'}. Don't miss out on this opportunity! {hashtags[0] if hashtags else '#Innovation'}"
-    ]
-    
-    # Dynamic product description
-    description = f"Introducing {product_name.title()}, the game-changing solution for {keywords[0] if keywords else 'modern needs'}. "
-    if len(keywords) > 1:
-        description += f"Our advanced approach to {keywords[1]} ensures you get the best {keywords[2] if len(keywords) > 2 else 'results'}. "
-    description += f"Whether you're looking for {keywords[0] if keywords else 'quality'} or {keywords[1] if len(keywords) > 1 else 'reliability'}, {product_name} delivers exceptional performance. "
-    description += f"Join the growing community of users who've discovered the power of {product_name} and see why it's becoming the preferred choice for {keywords[0] if keywords else 'success'}."
-    
-    # Dynamic CTAs
-    cta_ideas = [
-        f"Experience {product_name.title()} Today!",
-        f"Start Your {keywords[0].title() if keywords else 'Success'} Journey Now",
-        f"Get {product_name.title()} - Transform Your {keywords[1].title() if len(keywords) > 1 else 'Experience'}"
-    ]
-    
-    # Dynamic image prompts
-    image_prompts = [
-        f"Professional product photography of {product_name}, focused on {keywords[0] if keywords else 'quality'}, clean modern style",
-        f"Lifestyle image showing {product_name} in use, emphasizing {keywords[1] if len(keywords) > 1 else 'convenience'}, natural lighting",
-        f"Minimalist design featuring {product_name} with {keywords[2] if len(keywords) > 2 else 'elegant'} aesthetic, premium look"
-    ]
-    
-    return {
-        "taglines": tagline_templates,
-        "social_posts": social_posts,
-        "description": description,
-        "cta_ideas": cta_ideas,
-        "image_prompts": image_prompts,
-        "keywords": keywords
-    }
+def transcribe_audio_simple(audio_file):
+    """Simple audio transcription with progress tracking"""
+    try:
+        # Check if model is loaded
+        if st.session_state.whisper_model is None:
+            st.error("❌ Speech recognition model not loaded. Please try loading models first.")
+            return "Error: Speech recognition model not available"
+        
+        st.session_state.current_task = "Converting speech to text..."
+        
+        # Handle different input types
+        if isinstance(audio_file, str):
+            # File path
+            audio_input = audio_file
+        else:
+            # File-like object
+            audio_input = audio_file
+        
+        # Transcribe using pipeline
+        result = st.session_state.whisper_model(audio_input)
+        
+        st.session_state.current_task = ""
+        
+        # Handle different result formats
+        if isinstance(result, dict) and "text" in result:
+            return result["text"].strip()
+        elif isinstance(result, str):
+            return result.strip()
+        else:
+            return str(result).strip()
+        
+    except Exception as e:
+        st.session_state.current_task = ""
+        error_msg = f"Transcription error: {str(e)}"
+        st.error(error_msg)
+        
+        # Provide troubleshooting suggestions
+        if "librosa" in str(e).lower() or "soundfile" in str(e).lower():
+            st.error("🔧 Missing audio processing libraries. Install with:")
+            st.code("pip install librosa soundfile")
+        
+        return f"Error: {str(e)}"
 
-def generate_with_gemini(user_input):
-    """Generate enhanced content with Gemini"""
+def generate_content_with_gemini(prompt):
+    """Generate content using Gemini"""
+    if not GENAI_AVAILABLE:
+        return generate_content_offline(prompt)
+        
     try:
+        st.session_state.current_task = "Generating enhanced content with Gemini AI..."
+        
         model = genai.GenerativeModel('gemini-pro')
+        response = model.generate_content(f"""
+        Based on this input: "{prompt}"
+        
+        Create comprehensive marketing content with:
+        
+        ## Marketing Taglines
+        Generate 3 catchy, memorable taglines (max 12 words each)
+        
+        ## Social Media Posts  
+        Create 3 engaging social media posts (max 280 characters each)
         
-        prompt = f"""
-        Create detailed marketing content for: "{user_input}"
+        ## Product Description
+        Write 1 compelling product description (100-150 words)
         
-        Generate:
-        1. Three unique, catchy taglines (max 12 words each)
-        2. Three detailed social media posts (200-280 characters each) with relevant hashtags
-        3. One comprehensive product description (150-200 words)
-        4. Three compelling call-to-action phrases
-        5. Three specific image generation prompts
+        ## Image Generation Prompts
+        Provide 3 detailed prompts for AI image generation
         
-        Make everything highly relevant to the specific product/service described.
-        Format with clear markdown headers.
-        """
+        ## Call-to-Action Ideas
+        Suggest 3 effective call-to-action phrases
         
-        response = model.generate_content(prompt)
+        Format with clear markdown headers and numbered lists.
+        """)
+        
+        st.session_state.current_task = ""
         return response.text
         
     except Exception as e:
-        st.warning(f"Gemini AI unavailable: {e}")
-        return format_dynamic_content(generate_dynamic_content(user_input))
+        st.warning(f"Gemini error: {e}. Using offline generation.")
+        st.session_state.current_task = ""
+        return generate_content_offline(prompt)
 
-def format_dynamic_content(content):
-    """Format structured content for display"""
-    formatted = "## 🏷️ Marketing Taglines\n"
-    for i, tagline in enumerate(content["taglines"], 1):
-        formatted += f"{i}. **{tagline}**\n"
-    
-    formatted += "\n## 📱 Social Media Posts\n"
-    for i, post in enumerate(content["social_posts"], 1):
-        formatted += f"**Post {i}:**\n{post}\n\n"
+def generate_content_offline(prompt):
+    """Generate content using offline methods"""
+    st.session_state.current_task = "Generating content with offline templates..."
     
-    formatted += "## 📝 Product Description\n"
-    formatted += f"{content['description']}\n\n"
+    # Create structured content
+    content = {
+        "taglines": [
+            f"Experience {prompt} like never before",
+            f"Transform your world with {prompt}",
+            f"Discover the power of {prompt}"
+        ],
+        "social_posts": [
+            f"🌟 Ready to explore {prompt}? Join thousands who've already discovered the difference! #Innovation",
+            f"💫 {prompt} is changing the game! Don't miss out on this incredible opportunity. #GameChanger", 
+            f"🚀 The future of {prompt} is here! Experience what everyone's talking about. #FutureTech"
+        ],
+        "description": f"Discover the revolutionary world of {prompt}. Our innovative approach combines cutting-edge technology with user-friendly design to deliver an unmatched experience. Perfect for both beginners and experts, this solution transforms how you interact with {prompt}. Join thousands of satisfied users today!",
+        "image_prompts": [
+            f"Professional product photo of {prompt}, clean white background, studio lighting",
+            f"Modern minimalist illustration of {prompt}, flat design, vibrant colors",
+            f"Futuristic concept art of {prompt}, digital art, high quality, detailed"
+        ],
+        "cta_ideas": [
+            f"Get Started with {prompt} Today!",
+            f"Transform Your Experience Now",
+            f"Join the {prompt} Revolution"
+        ]
+    }
     
-    formatted += "## 🎯 Call-to-Action Ideas\n"
-    for i, cta in enumerate(content["cta_ideas"], 1):
-        formatted += f"{i}. {cta}\n"
+    # Format for display
+    formatted = format_content_display(content)
     
-    formatted += "\n## 🎨 Image Generation Prompts\n"
-    for i, prompt in enumerate(content["image_prompts"], 1):
-        formatted += f"{i}. {prompt}\n"
+    # Store both versions
+    st.session_state.generated_content['structured'] = content
+    st.session_state.current_task = ""
     
     return formatted
 
-def create_dynamic_flowchart(content_data, user_input):
-    """Create flowchart based on actual user content"""
+def create_flowchart_image(content_data):
+    """Create a simple flowchart visualization of the content"""
     try:
+        # Create image
         width, height = 800, 600
         image = Image.new('RGB', (width, height), 'white')
         draw = ImageDraw.Draw(image)
         
-        # Use default font
+        # Try to use a basic font, fall back to default if not available
         try:
-            font_title = ImageFont.truetype("arial.ttf", 18)
+            font_title = ImageFont.truetype("arial.ttf", 20)
             font_text = ImageFont.truetype("arial.ttf", 14)
             font_small = ImageFont.truetype("arial.ttf", 12)
         except:
-            font_title = font_text = font_small = ImageFont.load_default()
+            font_title = ImageFont.load_default()
+            font_text = ImageFont.load_default()
+            font_small = ImageFont.load_default()
         
         # Colors
-        colors = ["#2E86AB", "#A23B72", "#F18F01", "#C73E1D"]
-        
-        # Extract product name from input
-        words = user_input.split()
-        product = next((word for word in words if len(word) > 4), "Product")
-        
-        # Title with actual product
-        title = f"Marketing Strategy: {product.title()}"
-        draw.text((50, 20), title, fill="#333", font=font_title)
-        
-        # Dynamic sections based on content
-        sections = [
-            ("🏷️ Brand Message", content_data.get('taglines', ['Message'])[0][:40] + "..."),
-            ("📱 Social Strategy", content_data.get('social_posts', ['Strategy'])[0][:40] + "..."),
-            ("📝 Core Value", content_data.get('description', user_input)[:50] + "..."),
-            ("🎯 Action Plan", content_data.get('cta_ideas', ['Action'])[0])
-        ]
-        
-        # Draw sections
-        y = 80
-        for i, (title, content) in enumerate(sections):
-            color = colors[i % len(colors)]
-            
-            # Box
-            draw.rectangle([50, y, 750, y + 80], outline=color, width=2)
-            
-            # Title
-            draw.text((60, y + 10), title, fill=color, font=font_text)
-            
-            # Content
-            draw.text((60, y + 35), content, fill="#333", font=font_small)
-            
-            # Connect to next (except last)
-            if i < len(sections) - 1:
-                draw.line([(400, y + 80), (400, y + 100)], fill="#666", width=2)
-                draw.polygon([(395, y + 95), (405, y + 95), (400, y + 100)], fill="#666")
-            
-            y += 100
-        
-        # Add keywords if available
-        if 'keywords' in content_data and content_data['keywords']:
-            keywords_text = f"Key Terms: {', '.join(content_data['keywords'][:5])}"
-            draw.text((50, height - 50), keywords_text, fill="#666", font=font_small)
-        
-        draw.text((50, height - 30), f"Generated for: {user_input[:30]}...", fill="#999", font=font_small)
+        primary_color = "#2E86AB"
+        secondary_color = "#A23B72"
+        accent_color = "#F18F01"
+        text_color = "#333333"
+        
+        # Title
+        draw.text((width//2 - 150, 20), "Marketing Content Strategy", fill=text_color, font=font_title)
+        
+        # Draw boxes and content
+        y_offset = 80
+        box_height = 80
+        box_width = 180
+        
+        # Row 1: Taglines and Social Media
+        draw.rectangle([50, y_offset, 50 + box_width, y_offset + box_height], outline=primary_color, width=2)
+        draw.text((60, y_offset + 10), "🏷️ Taglines", fill=primary_color, font=font_text)
+        draw.text((60, y_offset + 35), f"• {content_data.get('taglines', ['Sample tagline'])[0][:25]}...", fill=text_color, font=font_small)
+        
+        draw.rectangle([width//2 + 50, y_offset, width//2 + 50 + box_width, y_offset + box_height], outline=secondary_color, width=2)
+        draw.text((width//2 + 60, y_offset + 10), "📱 Social Media", fill=secondary_color, font=font_text)
+        draw.text((width//2 + 60, y_offset + 35), f"• {content_data.get('social_posts', ['Sample post'])[0][:25]}...", fill=text_color, font=font_small)
+        
+        # Row 2: Description
+        y_offset += 120
+        draw.rectangle([width//4, y_offset, width*3//4, y_offset + box_height], outline=accent_color, width=2)
+        draw.text((width//4 + 10, y_offset + 10), "📝 Product Description", fill=accent_color, font=font_text)
+        desc_text = content_data.get('description', 'Product description goes here')[:50] + "..."
+        draw.text((width//4 + 10, y_offset + 35), desc_text, fill=text_color, font=font_small)
+        
+        # Row 3: CTAs and Image Ideas
+        y_offset += 120
+        draw.rectangle([50, y_offset, 50 + box_width, y_offset + box_height], outline=primary_color, width=2)
+        draw.text((60, y_offset + 10), "🎯 Call-to-Actions", fill=primary_color, font=font_text)
+        draw.text((60, y_offset + 35), f"• {content_data.get('cta_ideas', ['Sample CTA'])[0]}", fill=text_color, font=font_small)
+        
+        draw.rectangle([width//2 + 50, y_offset, width//2 + 50 + box_width, y_offset + box_height], outline=secondary_color, width=2)
+        draw.text((width//2 + 60, y_offset + 10), "🎨 Visual Ideas", fill=secondary_color, font=font_text)
+        draw.text((width//2 + 60, y_offset + 35), "• Professional photos", fill=text_color, font=font_small)
+        draw.text((width//2 + 60, y_offset + 50), "• Minimalist design", fill=text_color, font=font_small)
+        
+        # Draw connecting lines
+        draw.line([(width//2, 80 + box_height), (width//2, 200)], fill=text_color, width=2)
+        draw.line([(width//4 + box_width//2, 200 + box_height), (width//2, 320)], fill=text_color, width=2)
+        draw.line([(width*3//4 - box_width//2, 200 + box_height), (width//2, 320)], fill=text_color, width=2)
+        
+        # Add footer
+        draw.text((width//2 - 100, height - 30), "Generated by VoiceCanvas AI Studio", fill=text_color, font=font_small)
         
         return image
-    
+        
     except Exception as e:
-        st.error(f"Flowchart error: {e}")
+        st.error(f"Error creating flowchart: {e}")
         return None
 
+def format_content_display(content):
+    """Format content for nice display"""
+    if isinstance(content, dict):
+        formatted = ""
+        
+        if "taglines" in content:
+            formatted += "## 🏷️ Marketing Taglines\n"
+            for i, tagline in enumerate(content["taglines"], 1):
+                formatted += f"{i}. **{tagline}**\n"
+            formatted += "\n"
+        
+        if "social_posts" in content:
+            formatted += "## 📱 Social Media Posts\n"
+            for i, post in enumerate(content["social_posts"], 1):
+                formatted += f"**Post {i}:**\n{post}\n\n"
+        
+        if "description" in content:
+            formatted += "## 📝 Product Description\n"
+            formatted += f"{content['description']}\n\n"
+        
+        if "cta_ideas" in content:
+            formatted += "## 🎯 Call-to-Action Ideas\n"
+            for i, cta in enumerate(content["cta_ideas"], 1):
+                formatted += f"{i}. {cta}\n"
+            formatted += "\n"
+        
+        if "image_prompts" in content:
+            formatted += "## 🎨 Image Generation Prompts\n"
+            for i, prompt in enumerate(content["image_prompts"], 1):
+                formatted += f"{i}. {prompt}\n"
+        
+        return formatted
+    
+    return str(content)
+
+def handle_button_click(button_key):
+    """Handle button clicks to prevent multiple clicks"""
+    if not st.session_state.get(f'{button_key}_clicked', False):
+        st.session_state[f'{button_key}_clicked'] = True
+        return True
+    return False
+
+def reset_button_state(button_key):
+    """Reset button state"""
+    if f'{button_key}_clicked' in st.session_state:
+        st.session_state[f'{button_key}_clicked'] = False
+
 def main():
-    # Sidebar
+    # Sidebar with tips and status
     with st.sidebar:
         st.header("🎨 VoiceCanvas")
         st.markdown("*AI Content Studio*")
         
-        # Status
-        st.subheader("📊 Status")
+        # Load models button
+        if not st.session_state.models_loaded:
+            if st.button("🚀 Load AI Models", type="primary", use_container_width=True):
+                if handle_button_click("load_models"):
+                    with st.spinner("Loading AI models..."):
+                        success = load_models()
+                        reset_button_state("load_models")
+                        if success:
+                            st.rerun()
+        
+        # Status section
+        st.subheader("📊 System Status")
+        
+        gemini_available = setup_gemini()
         
-        gemini_status = setup_gemini()
+        col1, col2 = st.columns(2)
+        with col1:
+            st.metric("Mode", "Enhanced" if gemini_available else "Basic")
+        with col2:
+            st.metric("Status", "Ready" if not st.session_state.processing else "Working")
         
-        st.write(f"🤖 Speech Recognition: {'✅' if TRANSFORMERS_AVAILABLE else '❌'}")
-        st.write(f"🎙️ Audio Recording: {'✅' if AUDIO_REC_AVAILABLE else '❌'}")
-        st.write(f"🚀 Enhanced AI: {'✅' if gemini_status else '❌'}")
+        # Component status
+        st.write("🤖 **Components:**")
+        st.write(f"• Speech Recognition: {'✅' if st.session_state.models_loaded else '❌'}")
+        st.write(f"• Audio Recording: {'✅' if AUDIO_REC_AVAILABLE else '❌'}")
+        st.write(f"• Enhanced AI: {'✅' if gemini_available else '❌'}")
+        
+        # Current task indicator
+        if st.session_state.current_task:
+            st.info(f"🔄 {st.session_state.current_task}")
         
         st.markdown("---")
         
-        # Instructions
-        with st.expander("📖 How to Use"):
+        # Tips and help
+        st.subheader("💡 How to Use")
+        
+        with st.expander("🚀 Quick Start", expanded=True):
             st.markdown("""
-            1. **Input your idea** via voice, audio upload, or text
-            2. **Review and edit** your input
-            3. **Generate content** with one click
-            4. **Download** your marketing materials
+            1. **Load Models**: Click "Load AI Models" button first
+            2. **Input**: Use voice, upload audio, or type text
+            3. **Edit**: Review and refine your input
+            4. **Generate**: Create marketing content
+            5. **Visualize**: Generate flowchart of your strategy
+            6. **Export**: Download your materials
             """)
         
-        with st.expander("💡 Tips"):
+        with st.expander("🎯 Best Practices"):
             st.markdown("""
-            - Be specific about your product/service
+            **For Voice/Audio:**
+            - Speak clearly at normal pace
+            - Use quiet environment
+            - Describe your product/service
             - Mention target audience
-            - Include key features and benefits
-            - Use 30+ words for better results
+            
+            **For Text:**
+            - Be specific about features
+            - Include benefits and use cases
+            - Mention what makes it unique
+            - Use 50+ words for detail
+            """)
+        
+        with st.expander("⚙️ Setup (Optional)"):
+            st.markdown("""
+            **Enhanced Features:**
+            
+            Add environment variables:
+            - `GEMINI_API_KEY`: Advanced text generation
+            
+            **Get API Key:**
+            - [Google AI Studio](https://makersuite.google.com/app/apikey) (Free)
+            """)
+        
+        with st.expander("🛠️ Troubleshooting"):
+            st.markdown("""
+            **Common Issues:**
+            - "Speech recognition not available" → Click "Load AI Models"
+            - Audio processing errors → Install: `pip install librosa soundfile`
+            - Button not responding → Wait for processing to complete
+            - Slow processing → Models loading for first time
+            - Basic content only → Add GEMINI_API_KEY
             """)
 
     # Main content
     st.title("🎨 VoiceCanvas - AI Content Studio")
-    st.markdown("*Transform your ideas into marketing content*")
+    st.markdown("*Transform your ideas into comprehensive marketing content*")
+    
+    # Show model loading status
+    if not st.session_state.models_loaded:
+        st.warning("⚠️ AI models not loaded yet. Click 'Load AI Models' in the sidebar to enable speech recognition.")
     
-    # Input tabs
-    tabs = st.tabs(["🎙️ Record", "📁 Upload", "✍️ Type"] if AUDIO_REC_AVAILABLE else ["📁 Upload", "✍️ Type"])
+    # Main input area
+    st.header("💡 Share Your Idea")
     
-    # Recording tab
+    # Dynamic tabs based on available features
+    available_tabs = []
     if AUDIO_REC_AVAILABLE:
-        with tabs[0]:
-            st.info("🎤 Record your idea")
+        available_tabs.append("🎙️ Record")
+    available_tabs.extend(["📁 Upload", "✍️ Type"])
+    
+    tabs = st.tabs(available_tabs)
+    tab_index = 0
+    
+    # Recording tab (if available)
+    if AUDIO_REC_AVAILABLE:
+        with tabs[tab_index]:
+            st.info("🎤 Click the microphone button to start recording")
             
+            # Audio recorder
             wav_audio_data = st_audiorec()
             
-            if wav_audio_data and st.button("🔄 Transcribe", type="primary"):
-                with st.spinner("Converting speech to text..."):
-                    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
-                        tmp_file.write(wav_audio_data)
-                        st.session_state.transcription = transcribe_audio(tmp_file.name)
-                        os.unlink(tmp_file.name)
-                st.rerun()
+            if wav_audio_data is not None:
+                st.success("🎉 Audio recorded successfully!")
+                st.audio(wav_audio_data, format='audio/wav')
+                
+                col1, col2 = st.columns([1, 2])
+                with col1:
+                    if st.button("🔄 Transcribe Audio", key="transcribe_btn", type="primary"):
+                        if not st.session_state.models_loaded:
+                            st.error("Please load AI models first using the sidebar button.")
+                        else:
+                            if handle_button_click("transcribe"):
+                                st.session_state.processing = True
+                                # Process immediately
+                                with st.spinner("🎯 Converting your speech to text..."):
+                                    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
+                                        tmp_file.write(wav_audio_data)
+                                        transcription = transcribe_audio_simple(tmp_file.name)
+                                        st.session_state.transcription = transcription
+                                        os.unlink(tmp_file.name)
+                                st.session_state.processing = False
+                                reset_button_state("transcribe")
+                                st.rerun()
+                
+                with col2:
+                    if st.session_state.processing:
+                        st.info("🔄 Processing your audio...")
+        tab_index += 1
     
     # Upload tab
-    tab_idx = 1 if AUDIO_REC_AVAILABLE else 0
-    with tabs[tab_idx]:
-        st.info("📁 Upload audio file")
+    with tabs[tab_index]:
+        st.info("📁 Upload an audio file containing your idea")
         
-        uploaded_file = st.file_uploader("Choose audio file", type=['wav', 'mp3', 'm4a'])
+        uploaded_file = st.file_uploader(
+            "Choose audio file", 
+            type=['wav', 'mp3', 'm4a'],
+            help="Supported: WAV, MP3, M4A • Max 10MB • Best: 30 seconds or less"
+        )
         
-        if uploaded_file and st.button("🔄 Process Audio", type="primary"):
-            with st.spinner("Processing audio..."):
-                st.session_state.transcription = transcribe_audio(uploaded_file)
-            st.rerun()
+        if uploaded_file:
+            st.success("📄 File uploaded successfully!")
+            st.audio(uploaded_file)
+            
+            col1, col2 = st.columns([1, 2])
+            with col1:
+                if st.button("🔄 Process Audio", key="upload_transcribe", type="primary"):
+                    if not st.session_state.models_loaded:
+                        st.error("Please load AI models first using the sidebar button.")
+                    else:
+                        if handle_button_click("upload_process"):
+                            st.session_state.processing = True
+                            # Process immediately
+                            with st.spinner("🎯 Processing your audio file..."):
+                                transcription = transcribe_audio_simple(uploaded_file)
+                                st.session_state.transcription = transcription
+                            st.session_state.processing = False
+                            reset_button_state("upload_process")
+                            st.rerun()
+            
+            with col2:
+                if st.session_state.processing:
+                    st.info("🔄 Converting speech to text...")
+    
+    tab_index += 1
     
     # Text tab
-    tab_idx = 2 if AUDIO_REC_AVAILABLE else 1
-    with tabs[tab_idx]:
-        st.info("✍️ Describe your product/service")
+    with tabs[tab_index]:
+        st.info("✍️ Type or paste your product/service description")
         
         user_input = st.text_area(
-            "Your idea:",
-            placeholder="Example: A smart fitness tracker that monitors sleep, heart rate, and stress levels for busy professionals...",
-            height=120
+            "Describe your idea:",
+            placeholder="Example: A smart fitness tracker that monitors sleep patterns, heart rate, and stress levels. It provides personalized workout recommendations for busy professionals who want to maintain their health despite hectic schedules.",
+            height=150,
+            help="Be detailed! Include features, benefits, and target audience for best results."
         )
         
         if user_input:
             st.session_state.transcription = user_input
+            word_count = len(user_input.split())
+            
+            if word_count < 10:
+                st.warning("💡 Add more details for better results (at least 10 words)")
+            elif word_count < 30:
+                st.info("📝 Good start! Add more features/benefits for richer content")
+            else:
+                st.success(f"✅ Great detail! ({word_count} words)")
     
-    # Review and edit
+    # Show transcription and editing
     if st.session_state.transcription:
         st.markdown("---")
-        st.subheader("📝 Review Your Input")
+        st.header("📝 Review Your Input")
         
         edited_text = st.text_area(
-            "Edit your input:",
+            "Edit or refine your input:",
             value=st.session_state.transcription,
-            height=100
+            height=120,
+            key="edit_transcription",
+            help="Make any corrections or add more details"
         )
+        st.session_state.transcription = edited_text
         
-        # Generate content button
+        # Generate content section
+        st.markdown("---")
         col1, col2, col3 = st.columns([1, 2, 1])
+        
         with col2:
             if st.button("🚀 Generate Marketing Content", type="primary", use_container_width=True):
-                
-                with st.spinner("✨ Creating your marketing content..."):
-                    if setup_gemini():
-                        content = generate_with_gemini(edited_text)
-                        st.session_state.generated_content = {
-                            'text': content,
-                            'structured': generate_dynamic_content(edited_text)
-                        }
-                    else:
-                        structured_content = generate_dynamic_content(edited_text)
-                        formatted_content = format_dynamic_content(structured_content)
-                        st.session_state.generated_content = {
-                            'text': formatted_content,
-                            'structured': structured_content
-                        }
-                
-                st.success("✅ Content generated!")
-                st.rerun()
+                if handle_button_click("generate_content"):
+                    with st.spinner("✨ Creating comprehensive marketing content..."):
+                        if setup_gemini():
+                            content_text = generate_content_with_gemini(st.session_state.transcription)
+                            st.session_state.generated_content['text'] = content_text
+                        else:
+                            content_text = generate_content_offline(st.session_state.transcription)
+                            st.session_state.generated_content['text'] = content_text
+                    reset_button_state("generate_content")
+                    st.success("✅ Content generated successfully!")
+                    st.rerun()
     
-    # Display results
+    # Display generated content
     if st.session_state.generated_content:
         st.markdown("---")
         st.header("✨ Your Marketing Content")
         
-        # Show content
+        # Text content
         if 'text' in st.session_state.generated_content:
             st.markdown(st.session_state.generated_content['text'])
         
-        # Flowchart section
+        # Visual content section
         st.markdown("---")
-        st.subheader("📊 Strategy Visualization")
-        
-        if st.button("📊 Generate Strategy Flowchart", type="secondary"):
-            with st.spinner("Creating flowchart..."):
-                if 'structured' in st.session_state.generated_content:
-                    flowchart = create_dynamic_flowchart(
-                        st.session_state.generated_content['structured'], 
-                        st.session_state.transcription
-                    )
-                    if flowchart:
-                        st.session_state.generated_content['flowchart'] = flowchart
-            st.rerun()
+        st.subheader("🎨 Visual Content")
+        
+        col1, col2 = st.columns([1, 1])
+        
+        with col1:
+            if st.button("📊 Generate Strategy Flowchart", use_container_width=True, type="secondary"):
+                if handle_button_click("generate_flowchart"):
+                    with st.spinner("🎨 Creating strategy flowchart..."):
+                        if 'structured' in st.session_state.generated_content:
+                            flowchart_img = create_flowchart_image(st.session_state.generated_content['structured'])
+                            if flowchart_img:
+                                st.session_state.generated_content['flowchart'] = flowchart_img
+                        else:
+                            # Create basic flowchart from text content
+                            basic_data = {
+                                'taglines': ['Key message from your content'],
+                                'social_posts': ['Social media strategy'],
+                                'description': st.session_state.transcription[:100],
+                                'cta_ideas': ['Call to action'],
+                                'image_prompts': ['Visual elements']
+                            }
+                            flowchart_img = create_flowchart_image(basic_data)
+                            if flowchart_img:
+                                st.session_state.generated_content['flowchart'] = flowchart_img
+                    reset_button_state("generate_flowchart")
+                    st.success("📊 Flowchart created!")
+                    st.rerun()
+        
+        with col2:
+            st.info("💡 Generate a visual flowchart of your marketing strategy to better understand content relationships and flow.")
         
+        # Display generated flowchart
         if 'flowchart' in st.session_state.generated_content:
-            st.image(st.session_state.generated_content['flowchart'], use_column_width=True)
+            st.image(
+                st.session_state.generated_content['flowchart'], 
+                caption="Marketing Strategy Flowchart", 
+                use_column_width=True
+            )
         
-        # Download section
+        # Export section
         st.markdown("---")
-        st.subheader("📥 Download Content")
+        st.header("📥 Export Your Content")
         
         col1, col2, col3 = st.columns(3)
         
         with col1:
+            # Text export
             if 'text' in st.session_state.generated_content:
-                content_file = f"""VoiceCanvas Marketing Content
-Generated: {datetime.now().strftime('%Y-%m-%d %H:%M')}
-
-Input: {st.session_state.transcription}
+                content_export = f"""VOICECANVAS MARKETING CONTENT
+Generated: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
+Source: {st.session_state.transcription[:100]}...
 
 {st.session_state.generated_content['text']}
+
+---
+Created with VoiceCanvas AI Content Studio
 """
+                
                 st.download_button(
                     "📄 Download Text",
-                    content_file,
-                    f"marketing_content_{datetime.now().strftime('%Y%m%d_%H%M')}.txt",
-                    use_container_width=True
+                    content_export,
+                    file_name=f"marketing_content_{datetime.now().strftime('%Y%m%d_%H%M%S')}.txt",
+                    mime="text/plain",
+                    use_container_width=True,
+                    help="Download complete text content"
                 )
         
         with col2:
+            # JSON export
             if 'structured' in st.session_state.generated_content:
                 json_data = {
-                    "timestamp": datetime.now().isoformat(),
+                    "metadata": {
+                        "timestamp": datetime.now().isoformat(),
+                        "generator": "VoiceCanvas AI Studio",
+                        "mode": "Enhanced" if setup_gemini() else "Basic"
+                    },
                     "input": st.session_state.transcription,
                     "content": st.session_state.generated_content['structured']
                 }
+                
                 st.download_button(
                     "📊 Download Data",
                     json.dumps(json_data, indent=2),
-                    f"content_data_{datetime.now().strftime('%Y%m%d_%H%M')}.json",
+                    file_name=f"content_data_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json",
                     mime="application/json",
-                    use_container_width=True
+                    use_container_width=True,
+                    help="Download structured data (JSON)"
                 )
         
         with col3:
+            # Flowchart export
             if 'flowchart' in st.session_state.generated_content:
                 img_buffer = io.BytesIO()
                 st.session_state.generated_content['flowchart'].save(img_buffer, format="PNG")
+                
                 st.download_button(
                     "📊 Download Flowchart",
                     img_buffer.getvalue(),
-                    f"flowchart_{datetime.now().strftime('%Y%m%d_%H%M')}.png",
+                    file_name=f"strategy_flowchart_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png",
                     mime="image/png",
-                    use_container_width=True
+                    use_container_width=True,
+                    help="Download strategy flowchart"
                 )
+            else:
+                st.info("Generate flowchart first", icon="ℹ️")
+    
+    # Footer
+    st.markdown("---")
+    col1, col2, col3 = st.columns([1, 2, 1])
+    with col2:
+        st.markdown("🎨 **VoiceCanvas AI Content Studio**")
+        st.caption("Transform ideas into marketing magic • Built with Streamlit")
 
 if __name__ == "__main__":
     main()