Create app.py

app.py

@@ -0,0 +1,572 @@
+import os
+import sys
+import traceback
+import logging
+from pathlib import Path
+
+# Setup logging
+logging.basicConfig(
+    level=logging.INFO,
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
+)
+logger = logging.getLogger(__name__)
+
+# Add src to Python path
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), 'src'))
+
+# Import Gradio
+try:
+    import gradio as gr
+    logger.info("✓ Gradio imported successfully")
+except ImportError:
+    logger.error("✗ Gradio not found. Install with: pip install gradio")
+    sys.exit(1)
+
+# Import PyTorch
+try:
+    import torch
+    logger.info(f"✓ PyTorch imported successfully (GPU: {torch.cuda.is_available()})")
+except ImportError:
+    logger.error("✗ PyTorch not found. Install with: pip install torch")
+    sys.exit(1)
+
+# Import application components
+try:
+    from src.mayini_model import MAYINIModel, MAYINIVocabulary
+    from src.scraper import JobScraper
+    from src.customizer import ResumeCustomizer
+    from src.classifier import JobRelevanceClassifier
+    from src.agent import JobApplicationAgent
+    logger.info("✓ All application modules imported successfully")
+except ImportError as e:
+    logger.error(f"✗ Failed to import application modules: {e}")
+    logger.error("Make sure all files are in the 'src' directory")
+    traceback.print_exc()
+    sys.exit(1)
+
+# ============================================================================
+# INITIALIZATION
+# ============================================================================
+
+logger.info("=" * 70)
+logger.info("🤖 Initializing Job Application Agent (MAYINI Framework Edition)")
+logger.info("=" * 70)
+
+# Global variables to hold model instances
+mayini_model = None
+vocab = None
+customizer = None
+classifier = None
+scraper = None
+agent = None
+
+def initialize_components():
+    """Initialize all application components"""
+    global mayini_model, vocab, customizer, classifier, scraper, agent
+    
+    try:
+        logger.info("\n📦 Step 1: Initializing MAYINI Vocabulary...")
+        vocab = MAYINIVocabulary(vocab_size=5000)
+        logger.info("✓ MAYINI Vocabulary initialized (5000 tokens)")
+        
+        logger.info("\n📦 Step 2: Initializing MAYINI Model...")
+        mayini_model = MAYINIModel(
+            vocab_size=5000,
+            hidden_dim=256,
+            num_heads=8,
+            num_layers=4,
+            max_seq_len=512,
+            dropout=0.1
+        )
+        logger.info("✓ MAYINI Model initialized")
+        logger.info(f"  - Hidden Dimensions: 256")
+        logger.info(f"  - Attention Heads: 8")
+        logger.info(f"  - Transformer Layers: 4")
+        logger.info(f"  - Parameters: ~3.5M")
+        logger.info(f"  - Max Sequence Length: 512")
+        
+        # Set to evaluation mode
+        mayini_model.eval()
+        logger.info("✓ MAYINI Model set to evaluation mode")
+        
+        logger.info("\n📦 Step 3: Initializing Job Scraper...")
+        scraper = JobScraper()
+        logger.info("✓ Job Scraper initialized with sample jobs")
+        
+        logger.info("\n📦 Step 4: Initializing Resume Customizer...")
+        customizer = ResumeCustomizer(mayini_model, vocab)
+        logger.info("✓ Resume Customizer initialized")
+        
+        logger.info("\n📦 Step 5: Initializing Job Classifier...")
+        classifier = JobRelevanceClassifier()
+        classifier.mayini_model = mayini_model
+        classifier.mayini_vocab = vocab
+        logger.info("✓ Job Classifier initialized with MAYINI model")
+        
+        logger.info("\n📦 Step 6: Initializing Application Agent...")
+        agent = JobApplicationAgent(scraper, customizer, classifier)
+        logger.info("✓ Application Agent initialized")
+        
+        logger.info("\n" + "=" * 70)
+        logger.info("✅ ALL COMPONENTS INITIALIZED SUCCESSFULLY!")
+        logger.info("=" * 70 + "\n")
+        
+        return True
+        
+    except Exception as e:
+        logger.error(f"\n✗ INITIALIZATION ERROR: {e}")
+        logger.error(traceback.format_exc())
+        return False
+
+# Initialize on startup
+if not initialize_components():
+    logger.error("Failed to initialize components. Check errors above.")
+    sys.exit(1)
+
+# ============================================================================
+# INTERFACE FUNCTIONS
+# ============================================================================
+
+def search_jobs_interface(keywords: str, location: str, num_jobs: int) -> str:
+    """
+    Search and rank jobs based on keywords and location
+    Uses MAYINI embeddings for relevance matching
+    """
+    try:
+        if not keywords or not keywords.strip():
+            return "❌ Error: Please enter keywords"
+        
+        if not agent:
+            return "❌ Error: Application not initialized"
+        
+        logger.info(f"\n🔍 Searching for jobs: keywords='{keywords}', location='{location}', num={int(num_jobs)}")
+        
+        results = agent.search_and_apply(
+            keywords=keywords.strip(),
+            location=location.strip(),
+            num_jobs=int(num_jobs)
+        )
+        
+        if not results or 'applications' not in results:
+            return "❌ No jobs found matching your criteria"
+        
+        output = f"✅ **Search Results**\n"
+        output += f"Found: {results.get('total_jobs_found', 0)} jobs\n"
+        output += f"Relevant: {results.get('relevant_jobs', 0)} jobs\n"
+        output += f"Pass Rate: {results.get('pass_rate', 0):.1%}\n\n"
+        output += "---\n\n"
+        
+        for i, app in enumerate(results.get('applications', [])[:5], 1):
+            job = app.get('job', {})
+            score = app.get('relevance_score', 0)
+            
+            output += f"**{i}. {job.get('title', 'N/A')}**\n"
+            output += f"- Company: {job.get('company', 'N/A')}\n"
+            output += f"- Location: {job.get('location', 'N/A')}\n"
+            output += f"- 🎯 Relevance Score: **{score:.0%}**\n"
+            output += f"- 💰 Salary: {job.get('salary_range', 'Not specified')}\n"
+            output += f"- 📊 Experience Required: {job.get('experience_required', 'N/A')} years\n"
+            
+            match_details = app.get('match_details', {})
+            if match_details:
+                output += f"- ✓ Matching Skills: {', '.join(match_details.get('matching_skills', [])[:3])}\n"
+                output += f"- ✗ Missing Skills: {', '.join(match_details.get('missing_skills', [])[:2])}\n"
+            
+            output += "\n"
+        
+        logger.info(f"✓ Found and displayed {len(results.get('applications', [])[:5])} top results")
+        return output
+        
+    except Exception as e:
+        logger.error(f"Error in search_jobs_interface: {e}")
+        logger.error(traceback.format_exc())
+        return f"❌ Error: {str(e)}\n\nPlease try again with different inputs."
+
+def customize_resume_interface(job_title: str, company: str, requirements: str) -> str:
+    """
+    Customize resume for specific job using MAYINI embeddings
+    """
+    try:
+        if not job_title or not job_title.strip():
+            return "❌ Error: Please enter job title"
+        
+        if not customizer:
+            return "❌ Error: Customizer not initialized"
+        
+        logger.info(f"\n📄 Customizing resume for: {job_title} @ {company}")
+        
+        job = {
+            'title': job_title.strip(),
+            'company': company.strip(),
+            'description': f"{job_title} at {company}",
+            'requirements': [req.strip() for req in requirements.split(',') if req.strip()]
+        }
+        
+        customized = customizer.customize_for_job(job)
+        
+        output = f"✅ **Customized Resume**\n\n"
+        output += f"**Job:** {job_title} @ {company}\n\n"
+        
+        output += f"**Custom Summary:**\n"
+        output += f"{customized.get('summary', 'N/A')}\n\n"
+        
+        output += f"**Prioritized Skills:**\n"
+        skills = customized.get('skills', [])[:15]
+        for skill in skills:
+            output += f"• {skill}\n"
+        
+        if 'customized_for' in customized:
+            match_info = customized['customized_for']
+            output += f"\n**Match Information:**\n"
+            output += f"- Matching Skills: {len(match_info.get('matching_skills', []))} skills\n"
+            output += f"- Match Score: {match_info.get('match_score', 0):.0%}\n"
+        
+        logger.info(f"✓ Resume customized successfully")
+        return output
+        
+    except Exception as e:
+        logger.error(f"Error in customize_resume_interface: {e}")
+        logger.error(traceback.format_exc())
+        return f"❌ Error: {str(e)}\n\nPlease check your inputs and try again."
+
+def classify_job_interface(job_title: str, requirements: str) -> str:
+    """
+    Classify job relevance using MAYINI embeddings
+    """
+    try:
+        if not job_title or not job_title.strip():
+            return "❌ Error: Please enter job title"
+        
+        if not classifier:
+            return "❌ Error: Classifier not initialized"
+        
+        logger.info(f"\n🎯 Classifying job: {job_title}")
+        
+        job = {
+            'title': job_title.strip(),
+            'description': job_title.strip(),
+            'requirements': [req.strip() for req in requirements.split(',') if req.strip()] if requirements else [],
+            'location': 'Remote',
+            'company': 'Unknown',
+            'experience_required': 5,
+            'salary_range': 'Unknown'
+        }
+        
+        # Get sample resume skills
+        resume_skills = [
+            "Python", "Docker", "AWS", "PostgreSQL", "REST API",
+            "Microservices", "Git", "Kubernetes", "Machine Learning",
+            "Data Analysis", "SQL", "Linux", "Cloud Computing"
+        ]
+        
+        score = classifier.classify_job(job, resume_skills)
+        details = classifier.get_match_details(job, resume_skills)
+        
+        output = f"✅ **Job Classification Results**\n\n"
+        output += f"**Job Title:** {job_title}\n\n"
+        
+        # Relevance score
+        score_percent = score * 100
+        if score >= 0.8:
+            emoji = "🟢"
+            level = "EXCELLENT"
+        elif score >= 0.6:
+            emoji = "🟡"
+            level = "GOOD"
+        elif score >= 0.4:
+            emoji = "🟠"
+            level = "FAIR"
+        else:
+            emoji = "🔴"
+            level = "POOR"
+        
+        output += f"{emoji} **Relevance Score:** {score_percent:.1f}% ({level})\n"
+        output += f"**Recommendation:** {details.get('recommendation', 'N/A')}\n\n"
+        
+        # Decision
+        if score >= 0.6:
+            output += f"**Decision:** ✅ **APPLY**\n\n"
+        else:
+            output += f"**Decision:** ⏸️ **CONSIDER**\n\n"
+        
+        # Skill matching
+        output += f"**Matching Skills ({len(details.get('matching_skills', []))}):**\n"
+        for skill in details.get('matching_skills', []):
+            output += f"✓ {skill}\n"
+        
+        output += f"\n**Missing Skills ({len(details.get('missing_skills', []))}):**\n"
+        for skill in details.get('missing_skills', []):
+            output += f"✗ {skill}\n"
+        
+        logger.info(f"✓ Classification complete: {score_percent:.1f}%")
+        return output
+        
+    except Exception as e:
+        logger.error(f"Error in classify_job_interface: {e}")
+        logger.error(traceback.format_exc())
+        return f"❌ Error: {str(e)}\n\nPlease check your inputs and try again."
+
+def get_system_info() -> str:
+    """Get system information"""
+    try:
+        info = f"✅ **System Information**\n\n"
+        info += f"**Framework:** MAYINI Transformer Model\n"
+        info += f"**Vocabulary Size:** 5,000 tokens\n"
+        info += f"**Hidden Dimensions:** 256\n"
+        info += f"**Attention Heads:** 8\n"
+        info += f"**Transformer Layers:** 4\n"
+        info += f"**Total Parameters:** ~3.5M\n"
+        info += f"**Max Sequence Length:** 512\n\n"
+        
+        info += f"**Hardware:**\n"
+        info += f"- GPU Available: {torch.cuda.is_available()}\n"
+        if torch.cuda.is_available():
+            info += f"- GPU Name: {torch.cuda.get_device_name(0)}\n"
+        info += f"- PyTorch Version: {torch.__version__}\n\n"
+        
+        info += f"**Application Status:**\n"
+        info += f"- MAYINI Model: {'✓ Loaded' if mayini_model else '✗ Not Loaded'}\n"
+        info += f"- Vocabulary: {'✓ Loaded' if vocab else '✗ Not Loaded'}\n"
+        info += f"- Scraper: {'✓ Loaded' if scraper else '✗ Not Loaded'}\n"
+        info += f"- Customizer: {'✓ Loaded' if customizer else '✗ Not Loaded'}\n"
+        info += f"- Classifier: {'✓ Loaded' if classifier else '✗ Not Loaded'}\n"
+        info += f"- Agent: {'✓ Loaded' if agent else '✗ Not Loaded'}\n"
+        
+        return info
+    except Exception as e:
+        return f"❌ Error: {str(e)}"
+
+# ============================================================================
+# GRADIO INTERFACE
+# ============================================================================
+
+logger.info("\n🎨 Building Gradio Interface...")
+
+with gr.Blocks(
+    title="Job Application Agent - MAYINI Framework",
+    theme=gr.themes.Soft(),
+    css="""
+    .gradio-container { max-width: 1200px; margin: auto; }
+    .header { text-align: center; margin-bottom: 20px; }
+    .tab-content { padding: 20px; }
+    """
+) as demo:
+    
+    # Header
+    gr.Markdown("""
+    # 🤖 Job Application Agent
+    ### AI-Powered Job Search & Resume Customization
+    **Powered by MAYINI Framework** - A Custom Transformer-based ML Model
+    
+    > This application uses advanced machine learning to help you find and apply for the perfect job!
+    """)
+    
+    # Search & Match Jobs Tab
+    with gr.Tab("🔍 Search & Match Jobs"):
+        gr.Markdown("### Find jobs and get AI-powered relevance matching")
+        gr.Markdown("Enter your skills and find the best matching job opportunities using MAYINI embeddings.")
+        
+        with gr.Row():
+            with gr.Column():
+                search_keywords = gr.Textbox(
+                    label="Keywords",
+                    placeholder="python docker aws kubernetes",
+                    value="python",
+                    lines=2,
+                    info="Enter skills and technologies (comma-separated)"
+                )
+                search_location = gr.Textbox(
+                    label="Location",
+                    placeholder="Remote, San Francisco, New York",
+                    value="Remote",
+                    info="Enter job location"
+                )
+                search_num = gr.Slider(
+                    minimum=1,
+                    maximum=20,
+                    value=5,
+                    step=1,
+                    label="Number of Jobs",
+                    info="How many job results to display"
+                )
+                search_btn = gr.Button("🔍 Search Jobs", variant="primary", size="lg")
+            
+            with gr.Column():
+                search_output = gr.Markdown(
+                    value="### Results will appear here...",
+                    label="Search Results"
+                )
+        
+        search_btn.click(
+            fn=search_jobs_interface,
+            inputs=[search_keywords, search_location, search_num],
+            outputs=search_output,
+            show_progress=True
+        )
+    
+    # Customize Resume Tab
+    with gr.Tab("📄 Customize Resume"):
+        gr.Markdown("### Tailor your resume for specific job opportunities")
+        gr.Markdown("Get a customized resume summary and skill prioritization for any job posting.")
+        
+        with gr.Row():
+            with gr.Column():
+                customize_title = gr.Textbox(
+                    label="Job Title",
+                    placeholder="Senior Python Developer",
+                    info="The position you're interested in"
+                )
+                customize_company = gr.Textbox(
+                    label="Company Name",
+                    placeholder="Tech Giants Inc",
+                    info="Company name (optional)"
+                )
+                customize_req = gr.Textbox(
+                    label="Job Requirements",
+                    placeholder="Python, Docker, AWS, Kubernetes, PostgreSQL",
+                    lines=3,
+                    info="List requirements (comma-separated)"
+                )
+                customize_btn = gr.Button("✨ Customize Resume", variant="primary", size="lg")
+            
+            with gr.Column():
+                customize_output = gr.Markdown(
+                    value="### Customized resume will appear here...",
+                    label="Customized Resume"
+                )
+        
+        customize_btn.click(
+            fn=customize_resume_interface,
+            inputs=[customize_title, customize_company, customize_req],
+            outputs=customize_output,
+            show_progress=True
+        )
+    
+    # Classify Job Tab
+    with gr.Tab("🎯 Classify Job Relevance"):
+        gr.Markdown("### Check how relevant a job is to your skills")
+        gr.Markdown("Get a detailed analysis of job relevance with matching and missing skills.")
+        
+        with gr.Row():
+            with gr.Column():
+                classify_title = gr.Textbox(
+                    label="Job Title",
+                    placeholder="Machine Learning Engineer",
+                    info="The job position to classify"
+                )
+                classify_req = gr.Textbox(
+                    label="Job Requirements",
+                    placeholder="Python, PyTorch, TensorFlow, Machine Learning, SQL",
+                    lines=3,
+                    info="Required skills (comma-separated)"
+                )
+                classify_btn = gr.Button("🎯 Classify Job", variant="primary", size="lg")
+            
+            with gr.Column():
+                classify_output = gr.Markdown(
+                    value="### Classification results will appear here...",
+                    label="Classification Results"
+                )
+        
+        classify_btn.click(
+            fn=classify_job_interface,
+            inputs=[classify_title, classify_req],
+            outputs=classify_output,
+            show_progress=True
+        )
+    
+    # System Info Tab
+    with gr.Tab("ℹ️ About & System Info"):
+        gr.Markdown("""
+        ## About Job Application Agent
+        
+        This application uses the **MAYINI Framework** - a custom Transformer-based 
+        machine learning model specifically designed for semantic job matching and resume customization.
+        
+        ### Key Features
+        
+        🔍 **Intelligent Job Search**
+        - AI-powered job discovery and ranking
+        - Real-time relevance scoring
+        - Multi-criteria filtering
+        
+        📄 **Resume Customization**
+        - Automatic resume tailoring
+        - Skill prioritization
+        - Semantic matching with MAYINI embeddings
+        
+        🎯 **Job Classification**
+        - Relevance scoring (0-100%)
+        - Skill gap analysis
+        - Recommendation engine
+        
+        ### MAYINI Framework Details
+        
+        **Architecture:**
+        - Model Type: Transformer Encoder
+        - Vocabulary Size: 5,000 tokens
+        - Hidden Dimensions: 256
+        - Attention Heads: 8
+        - Transformer Layers: 4
+        - Total Parameters: ~3.5 million
+        - Max Sequence Length: 512 tokens
+        
+        **Capabilities:**
+        - Text tokenization and encoding
+        - 256-dimensional embeddings
+        - Multi-head self-attention
+        - Semantic understanding
+        - Job-resume similarity matching
+        
+        ### Technology Stack
+        - **ML Framework**: MAYINI Transformer (Custom)
+        - **Interface**: Gradio 4.0+
+        - **Deep Learning**: PyTorch 2.0+
+        - **Language**: Python 3.8+
+        - **Deployment**: Hugging Face Spaces
+        
+        ### Repository & Links
+        - **GitHub**: [907-bot/Job-Application-Agent](https://github.com/907-bot/Job-Application-Agent)
+        - **License**: Apache 2.0
+        - **Status**: Production Ready
+        
+        ---
+        """)
+        
+        info_btn = gr.Button("🖥️ Show System Info", variant="primary")
+        info_output = gr.Markdown()
+        
+        info_btn.click(fn=get_system_info, outputs=info_output)
+
+logger.info("✓ Gradio Interface built successfully\n")
+
+# ============================================================================
+# LAUNCH APPLICATION
+# ============================================================================
+
+if __name__ == "__main__":
+    logger.info("=" * 70)
+    logger.info("🚀 LAUNCHING JOB APPLICATION AGENT")
+    logger.info("=" * 70)
+    
+    logger.info("\n📍 Access the application at:")
+    logger.info("   Local: http://127.0.0.1:7860")
+    logger.info("   Hugging Face Spaces: Check your Space URL")
+    logger.info("\n💡 Features:")
+    logger.info("   1. Search & Match Jobs")
+    logger.info("   2. Customize Resume")
+    logger.info("   3. Classify Job Relevance")
+    logger.info("   4. System Information\n")
+    
+    try:
+        demo.queue(max_size=32, concurrency_count=4).launch(
+            server_name="0.0.0.0",
+            server_port=7860,
+            show_error=True,
+            share=False,
+            show_api=True
+        )
+    except Exception as e:
+        logger.error(f"Failed to launch: {e}")
+        logger.error(traceback.format_exc())
+        sys.exit(1)