app.py

import streamlit as st
import json
from typing import Dict, List, Any
import re

def format_project_response(project: dict, indent_level: int = 0) -> str:
    """Format project details with proper indentation and spacing"""
    indent = "  " * indent_level
    response = [f"\n{indent}• {project['name']}:"]
    
    # Add description with proper indentation
    description_lines = project['description'].split('. ')
    response.extend([f"{indent}  {line.strip()}." for line in description_lines])
    
    # Add technologies with proper line break
    if 'skills_used' in project:
        response.append(f"\n{indent}  Technologies: {', '.join(project['skills_used'])}")
    
    # Add status and notes
    if 'status' in project:
        status = project['status']
        if 'development' in status.lower() or 'progress' in status.lower():
            response.append(f"\n{indent}  Status: {status}")
            if 'confidentiality_note' in project:
                response.append(f"{indent}  Note: {project['confidentiality_note']}")
    
    return '\n'.join(response) + '\n'

def analyze_job_requirements(text: str, knowledge_base: dict) -> Dict[str, List[str]]:
    """Analyze job requirements and match with skills"""
    text_lower = text.lower()
    
    # Common ML/AI related keywords
    tech_keywords = {
        'machine learning': ['ml', 'machine learning', 'deep learning', 'neural networks'],
        'data science': ['data science', 'data analysis', 'analytics', 'statistics'],
        'programming': ['python', 'sql', 'programming', 'coding'],
        'tools': ['tableau', 'powerbi', 'visualization', 'git'],
        'cloud': ['aws', 'azure', 'cloud', 'deployment']
    }
    
    # Extract matches from knowledge base
    matches = {category: [] for category in tech_keywords}
    my_skills = {
        skill.lower() 
        for skill_type in knowledge_base['skills']['technical_skills'].values() 
        for skill_list in skill_type.values() 
        for skill in skill_list
    }
    
    # Find matching skills in each category
    for category, keywords in tech_keywords.items():
        for keyword in keywords:
            if keyword in text_lower and any(skill in keyword or keyword in skill for skill in my_skills):
                matches[category].append(keyword)
    
    return matches

def handle_perspective_query(query: str, knowledge_base: dict) -> str:
    """Handle philosophical or perspective-based queries"""
    query_lower = query.lower()
    perspectives = knowledge_base.get('perspectives', {})
    
    # Market-related queries
    if any(word in query_lower for word in ['market', 'opportunity', 'job', 'hiring']):
        if any(word in query_lower for word in ['down', 'bad', 'difficult', 'tough']):
            response_parts = [
                "• My Perspective on the Current Market:",
                f"  {perspectives['market_outlook']['job_market']}",
                "",
                "• My Strategic Approach:",
                f"  {perspectives['market_outlook']['strategy']}",
                "",
                "• My Unique Value Proposition:",
                f"  {perspectives['market_outlook']['value_proposition']}"
            ]
            return '\n'.join(response_parts)
    
    # Learning and growth queries
    elif any(word in query_lower for word in ['learn', 'study', 'growth']):
        return f"• My Learning Philosophy:\n  {perspectives['learning_philosophy']}"
    
    # Handle non-portfolio queries gracefully
    return knowledge_base['common_queries']['general']

def generate_response(query: str, knowledge_base: dict) -> str:
    """Generate enhanced responses using the knowledge base"""
    query_lower = query.lower()
    
    # Handle project listing requests
    if any(word in query_lower for word in ['list', 'project', 'portfolio', 'built', 'created', 'developed']):
        response_parts = ["Here are my key projects:"]
        
        # Major Projects (under development)
        response_parts.append("\nMajor Projects (In Development):")
        for project in knowledge_base['projects']['major_projects']:
            response_parts.append(format_project_response(project, indent_level=1))
        
        # Algorithm Implementation Projects
        response_parts.append("\nCompleted Algorithm Implementation Projects:")
        for project in knowledge_base['projects']['algorithm_practice_projects']:
            response_parts.append(format_project_response(project, indent_level=1))
        
        response = '\n'.join(response_parts)
        return add_relevant_links(response, query, knowledge_base)
    
    # Handle job description analysis
    elif len(query.split()) > 20 and any(phrase in query_lower for phrase in 
        ['requirements', 'qualifications', 'looking for', 'job description']):
        
        matches = analyze_job_requirements(query, knowledge_base)
        relevant_projects = find_relevant_projects(query, knowledge_base['projects']['major_projects'])
        
        response_parts = ["Based on the job requirements, here's how my profile aligns:\n"]
        
        # Technical Skills Match
        if any(matches.values()):
            response_parts.append("• Technical Skills Alignment:")
            for category, skills in matches.items():
                if skills:
                    response_parts.append(f"  - Strong {category} skills: {', '.join(skills)}")
            response_parts.append("")
        
        # Project Experience
        if relevant_projects:
            response_parts.append("• Relevant Project Experience:")
            for project in relevant_projects:
                desc = f"  - {project['name']}: {project['description']}"
                response_parts.append(desc)
            response_parts.append("")
        
        # Education and Background
        response_parts.extend([
            "• Education and Background:",
            "  - Advanced AI/ML education in Canada",
            "  - Unique commerce background providing business perspective",
            "  - Strong foundation in practical ML implementation",
            ""
        ])
        
        return '\n'.join(response_parts)
    
    # Handle perspective/philosophical queries
    elif any(word in query_lower for word in ['market', 'think', 'believe', 'opinion', 'weather']):
        return handle_perspective_query(query, knowledge_base)
    
    # Handle story/background queries
    elif any(word in query_lower for word in ['background', 'journey', 'story', 'transition']):
        return format_story_response(knowledge_base)
    
    # Default response
    return format_default_response(knowledge_base)

def main():
    st.title("🤖 Meet Rini - AI-Powered Insights on Manyue's World")
    
    # Initialize session state
    if "messages" not in st.session_state:
        st.session_state.messages = []
    if "knowledge_base" not in st.session_state:
        try:
            with open('knowledge_base.json', 'r', encoding='utf-8') as f:
                st.session_state.knowledge_base = json.load(f)
        except FileNotFoundError:
            st.error("Knowledge base file not found.")
            return
    
    # Display welcome message
    if "displayed_welcome" not in st.session_state:
        st.write("""
        Hi there! I'm Rini, Manyue's AI-powered assistant. I'm here to represent Manyue and share insights about:
        - My journey from commerce to ML/AI
        - My technical skills and projects
        - My fit for ML/AI roles
        - My perspective on the tech industry
        - You can also paste job descriptions to see how my profile matches!
        """)
        st.session_state.displayed_welcome = True

    # Create two columns with chat history in scrollable container
    col1, col2 = st.columns([3, 1])
    
    with col1:
        # Chat container for better scrolling
        chat_container = st.container()
        with chat_container:
            for message in st.session_state.messages:
                with st.chat_message(message["role"]):
                    st.markdown(message["content"])
        
        # Chat input at bottom
        if prompt := st.chat_input("Ask me anything or paste a job description..."):
            # Add user message
            st.session_state.messages.append({"role": "user", "content": prompt})
            
            try:
                # Generate and display response
                with st.chat_message("assistant"):
                    response = generate_response(prompt, st.session_state.knowledge_base)
                    st.markdown(response)
                    st.session_state.messages.append({"role": "assistant", "content": response})
            except Exception as e:
                st.error(f"An error occurred: {str(e)}")
            
            st.rerun()
    
    with col2:
        st.subheader("Quick Questions")
        example_questions = [
            "Tell me about your ML projects",
            "What are your technical skills?",
            "What makes you stand out?",
            "What's your journey into ML?",
            "Your view on the current market?"
        ]
        
        for question in example_questions:
            if st.button(question):
                st.session_state.messages.append({"role": "user", "content": question})
                st.rerun()
        
        st.markdown("---")
        if st.button("Clear Chat"):
            st.session_state.messages = []
            st.rerun()

if __name__ == "__main__":
    main()