Update src/streamlit_app.py

src/streamlit_app.py

@@ -1,556 +1,109 @@
-# Personal Finance Chatbot: AI-Powered Financial Advisor
-# Enhanced with OpenAI GPT for superior accuracy and intelligence
-
-import os
-import io
-import re
-import json
-from dataclasses import dataclass
-from typing import List, Dict, Optional, Any, Union
-
-import pandas as pd
 import streamlit as st
+from transformers import pipeline
 
-# OpenAI Integration
-try:
-    from openai import OpenAI
-    OPENAI_AVAILABLE = True
-except ImportError:
-    OPENAI_AVAILABLE = False
-
-# HuggingFace Fallback
-try:
-    from transformers import pipeline
-    HF_AVAILABLE = True
-    HF_PIPELINE = pipeline
-except ImportError:
-    HF_AVAILABLE = False
-    HF_PIPELINE = None
-
-st.set_page_config(
-    page_title="FinanceAI - Smart Financial Advisor", 
-    page_icon="🤖", 
-    layout="wide",
-    initial_sidebar_state="collapsed"
-)
-
-if "chat_history" not in st.session_state:
-    st.session_state.chat_history = []
-if "provider_inited" not in st.session_state:
-    st.session_state.provider_inited = False
-if "provider_name" not in st.session_state:
-    st.session_state.provider_name = "openai"
-
-class AIProvider:
-    def __init__(self):
-        self.name = "base"
-    def generate(self, prompt: str, max_tokens: int = 512) -> str:
-        raise NotImplementedError
-
-class OpenAIProvider(AIProvider):
-    def __init__(self):
-        super().__init__()
-        self.name = "OpenAI GPT-4"
-        self.client = None
-        api_key = os.environ.get("OPENAI_API_KEY")
-        if OPENAI_AVAILABLE and api_key:
-            try:
-                self.client = OpenAI(api_key=api_key)
-                # Test the connection
-                self.client.models.list()
-                self.available = True
-            except Exception as e:
-                st.warning(f"OpenAI connection failed: {e}")
-                self.available = False
-                self.client = None
-        else:
-            self.available = False
-            
-    def generate(self, prompt: str, max_tokens: int = 768) -> str:
-        if not self.available or not self.client:
-            return "OpenAI service unavailable. Please check your API key."
-        
-        try:
-            # Use GPT-4 for superior financial advice
-            response = self.client.chat.completions.create(
-                model="gpt-4o-mini",  # Using the latest efficient model
-                messages=[
-                    {"role": "system", "content": "You are FinanceAI, a highly knowledgeable financial advisor specializing in Indian markets, tax laws, and investment strategies. Provide accurate, actionable advice with specific amounts and calculations."},
-                    {"role": "user", "content": prompt}
-                ],
-                max_tokens=max_tokens,
-                temperature=0.7,
-                top_p=0.9
-            )
-            return response.choices[0].message.content.strip()
-        except Exception as e:
-            return f"Error generating response: {str(e)}"
-
-class HuggingFaceProvider(AIProvider):
-    def __init__(self):
-        super().__init__()
-        self.name = "HuggingFace Fallback"
-        self.gen = None
-        if HF_AVAILABLE and HF_PIPELINE:
-            try:
-                self.gen = HF_PIPELINE("text2text-generation", model="google/flan-t5-base")
-                self.available = True
-            except Exception as e:
-                st.warning(f"HuggingFace pipeline failed to load: {e}")
-                self.available = False
-        else:
-            self.available = False
-            
-    def generate(self, prompt: str, max_tokens: int = 512) -> str:
-        if not self.available or self.gen is None:
-            return self._fallback_response(prompt)
-        
-        try:
-            out = self.gen(prompt, max_length=min(1024, max_tokens), do_sample=False)
-            return out[0]["generated_text"].strip()
-        except Exception:
-            return self._fallback_response(prompt)
-    
-    def _fallback_response(self, prompt: str) -> str:
-        return (
-            "🤖 **AI-Enhanced Financial Guidance**\n\n"
-            "Based on your query, here are key recommendations:\n"
-            "• **Emergency Fund**: Build 6 months of expenses in liquid funds\n"
-            "• **Tax Saving**: Utilize Section 80C (₹1.5L limit) through ELSS/PPF\n"
-            "• **Investment**: Balance equity (60%) and debt (40%) based on risk profile\n"
-            "• **Insurance**: Term life insurance of 10x annual income\n\n"
-            "For personalized calculations, please provide your monthly income and goals."
+# ---- Load HuggingFace Model ----
+@st.cache_resource
+def load_model():
+    try:
+        # Try Falcon-7B first (best results, requires GPU)
+        return pipeline(
+            "text-generation",
+            model="tiiuae/falcon-7b-instruct",
+            device_map="auto",
+            trust_remote_code=True
         )
-
-CATEGORIES = {
-    "groceries": ["grocery", "supermarket", "food", "mart", "vegetables", "fruits"],
-    "rent": ["rent", "landlord", "housing"],
-    "utilities": ["electric", "water", "gas", "utility", "internet", "mobile", "phone"],
-    "transport": ["uber", "ola", "fuel", "bus", "metro", "train", "cab", "petrol", "diesel"],
-    "entertainment": ["netflix", "spotify", "movie", "cinema", "concert", "game", "youtube"],
-    "health": ["pharmacy", "doctor", "hospital", "clinic", "medicine", "health"],
-    "eating_out": ["restaurant", "cafe", "bar", "eatery", "diner", "zomato", "swiggy"],
-    "shopping": ["amazon", "flipkart", "myntra", "shop", "store", "clothing"],
-    "income": ["salary", "stipend", "bonus", "interest", "dividend", "freelance"],
-    "investment": ["sip", "mutual fund", "stock", "gold", "ppf", "fd"],
-    "insurance": ["insurance", "premium", "policy"],
-    "education": ["fees", "books", "course", "training"],
-    "travel": ["flight", "hotel", "vacation", "trip"]
-}
-
-def categorize(desc: str) -> str:
-    desc_l = (desc or "").lower()
-    for cat, keys in CATEGORIES.items():
-        if any(k in desc_l for k in keys):
-            return cat
-    return "other"
-
-@dataclass
-class UserProfile:
-    name: str
-    user_type: str
-    age: int
-    country: str
-    monthly_income: float
-    risk_tolerance: str
-    goals: str
-    
-    def get_risk_level(self) -> str:
-        risk_map = {
-            "conservative": "low",
-            "moderate": "medium", 
-            "aggressive": "high"
-        }
-        return risk_map.get(self.risk_tolerance.lower(), "medium")
-
-def load_transactions(uploaded_file: Optional[io.BytesIO]) -> pd.DataFrame:
-    if uploaded_file is None:
-        # Enhanced sample data with more realistic transactions
-        data = {
-            "date": pd.date_range("2025-01-01", periods=30, freq="D"),
-            "description": [
-                "Salary Credit", "Rent Payment", "Grocery Store", "Restaurant Bill", "Metro Card", 
-                "Internet Bill", "Pharmacy", "Movie Ticket", "Amazon Purchase", "Fuel Station",
-                "Bonus Credit", "Electricity Bill", "Coffee Shop", "Supermarket", "Hospital",
-                "Netflix Subscription", "Uber Ride", "Water Bill", "Gym Membership", "Flipkart",
-                "Bus Pass", "Medicine", "Dividend Credit", "Train Ticket", "SIP Investment",
-                "Insurance Premium", "Mobile Recharge", "Zomato Order", "Book Purchase", "ATM Withdrawal"
-            ],
-            "amount": [
-                75000, -18000, -3200, -1200, -500, -1500, -800, -400, -2800, -2000,
-                15000, -2200, -300, -2800, -3500, -800, -250, -600, -2000, -1500,
-                -400, -450, 2500, -180, -5000, -3000, -399, -850, -550, -2000
-            ],
-        }
-        df = pd.DataFrame(data)
+    except Exception as e1:
+        st.warning("⚠️ Falcon-7B could not load. Falling back to FLAN-T5-Base (lighter).")
+        try:
+            return pipeline("text2text-generation", model="google/flan-t5-base")
+        except Exception as e2:
+            st.error(f"❌ Failed to load any model.\nFalcon error: {e1}\nFLAN error: {e2}")
+            return None
+
+generator = load_model()
+
+# ---- Build prompt ----
+def build_prompt(user_input, profile, summary, is_t5=False):
+    profile_text = (
+        f"You are FinanceAI, an expert Indian financial advisor.\n\n"
+        f"Profile:\n"
+        f"Age: {profile['age']}, "
+        f"Occupation: {profile['occupation']}, "
+        f"Income: ₹{profile['income']}/month, "
+        f"Risk Appetite: {profile['risk_appetite']}, "
+        f"Goals: {', '.join(profile['goals'])}\n\n"
+    )
+    summary_text = f"Financial Summary:\n{summary}\n\n" if summary else ""
+    if is_t5:
+        return f"{profile_text}{summary_text}Question: {user_input}\nAnswer with step-by-step financial advice."
     else:
-        df = pd.read_csv(uploaded_file)
-    
-    df["category"] = df["description"].apply(categorize)
-    return df
-
-def budget_summary(df: pd.DataFrame, monthly_income_hint: Optional[float] = None) -> Dict[str, Union[float, str]]:
-    try:
-        df["month"] = pd.to_datetime(df["date"]).dt.to_period("M").astype(str)
-        income = df.loc[df["amount"] > 0, "amount"].sum()
-        expenses = -df.loc[df["amount"] < 0, "amount"].sum()
-        net = income - expenses
-        
-        if monthly_income_hint and monthly_income_hint > 0:
-            income = max(income, monthly_income_hint)
-            net = income - expenses
-            
-        savings_rate = (net / income) * 100 if income > 0 else 0.0
-        
-        # Category breakdown
-        expense_by_category = df[df["amount"] < 0].groupby("category")["amount"].sum().abs()
-        top_expenses = expense_by_category.nlargest(5)
-        
-        return {
-            "income_total": float(round(income, 2)),
-            "expense_total": float(round(expenses, 2)),
-            "net_savings": float(round(net, 2)),
-            "savings_rate_pct": float(round(savings_rate, 2)),
-            "top_expenses": top_expenses.to_dict(),
-            "expense_breakdown": expense_by_category.to_dict()
+        return f"{profile_text}{summary_text}User question: {user_input}\n\nGive clear, step-by-step financial advice with examples."
+
+# ---- Streamlit UI ----
+def main():
+    st.set_page_config(page_title="FinanceAI Chatbot", layout="wide")
+    st.title("💰 Personal Finance Chatbot")
+    st.write("This chatbot runs on **Hugging Face Spaces**.\n\n"
+             "It tries **Falcon-7B-Instruct** (GPU) first, then falls back to **FLAN-T5-Base** (CPU-friendly).")
+
+    # --- Initialize session state ---
+    if "profile" not in st.session_state:
+        st.session_state.profile = {
+            "age": 25,
+            "occupation": "Student",
+            "income": 50000,
+            "risk_appetite": "Moderate",
+            "goals": [],
         }
-    except Exception as e:
-        return {
-            "income_total": monthly_income_hint or 50000,
-            "expense_total": 35000,
-            "net_savings": 15000,
-            "savings_rate_pct": 30.0,
-            "top_expenses": {},
-            "expense_breakdown": {}
-        }
-
-def generate_ai_financial_plan(profile: UserProfile, summary: Dict[str, Union[float, str]]) -> str:
-    """Generate comprehensive AI-powered financial plan"""
-    
-    income = summary["income_total"]
-    savings = summary["net_savings"]
-    savings_rate = summary["savings_rate_pct"]
-    
-    plan = []
-    
-    # Emergency Fund Strategy
-    emergency_fund_target = profile.monthly_income * 6
-    monthly_emergency_save = min(profile.monthly_income * 0.15, 10000)
-    plan.append(f"🚨 **Emergency Fund Strategy**")
-    plan.append(f"Target: ₹{emergency_fund_target:,.0f} (6 months expenses)")
-    plan.append(f"Monthly allocation: ₹{monthly_emergency_save:,.0f}")
-    plan.append(f"Investment: Liquid funds or high-yield savings (6-8% returns)")
-    
-    # Tax Optimization
-    plan.append(f"\n📊 **Tax Optimization Plan**")
-    max_80c = 150000
-    recommended_80c = min(profile.monthly_income * 12 * 0.125, max_80c)
-    monthly_80c = recommended_80c / 12
-    plan.append(f"Section 80C: ₹{monthly_80c:,.0f}/month (₹{recommended_80c:,.0f}/year)")
-    plan.append(f"Options: ELSS (12-15% returns), PPF (7.1% tax-free), NSC")
-    
-    # Investment Strategy
-    plan.append(f"\n🚀 **Investment Allocation Strategy**")
-    investable_amount = max(0, savings - monthly_emergency_save)
-    
-    if investable_amount > 1000:
-        risk_level = profile.get_risk_level()
-        if risk_level == "low":
-            equity_pct, debt_pct = 40, 60
-            expected_return = "8-10%"
-        elif risk_level == "high":
-            equity_pct, debt_pct = 80, 20
-            expected_return = "12-15%"
+    if "summary" not in st.session_state:
+        st.session_state.summary = "Income ₹50,000; Expenses ₹30,000; Savings ₹20,000"
+    if "chat_history" not in st.session_state:
+        st.session_state.chat_history = []
+
+    # Sidebar profile setup
+    st.sidebar.header("👤 User Profile")
+    st.session_state.profile["age"] = st.sidebar.number_input("Age", min_value=18, max_value=100, value=st.session_state.profile["age"])
+    st.session_state.profile["occupation"] = st.sidebar.selectbox("Occupation", ["Student", "Salaried", "Freelancer", "Business Owner"], index=["Student", "Salaried", "Freelancer", "Business Owner"].index(st.session_state.profile["occupation"]))
+    st.session_state.profile["income"] = st.sidebar.number_input("Monthly Income (₹)", min_value=0, value=st.session_state.profile["income"])
+    st.session_state.profile["risk_appetite"] = st.sidebar.selectbox("Risk Appetite", ["Low", "Moderate", "High"], index=["Low", "Moderate", "High"].index(st.session_state.profile["risk_appetite"]))
+    st.session_state.profile["goals"] = st.sidebar.multiselect("Financial Goals", ["Emergency Fund", "Retirement", "Travel", "Home", "Education", "Wealth Growth"], default=st.session_state.profile["goals"])
+
+    st.session_state.summary = st.sidebar.text_area("📊 Financial Summary", st.session_state.summary)
+
+    # Chat section
+    st.header("💬 Chat with FinanceAI")
+
+    user_input = st.text_input("Ask your financial question:")
+
+    if st.button("Send") and user_input:
+        if generator:
+            is_t5 = "text2text" in str(type(generator))
+            prompt = build_prompt(user_input, st.session_state.profile, st.session_state.summary, is_t5=is_t5)
+
+            if is_t5:
+                response = generator(prompt, max_length=256)
+                ai_response = response[0]['generated_text']
+            else:
+                response = generator(
+                    prompt,
+                    max_length=512,
+                    do_sample=True,
+                    top_p=0.9,
+                    temperature=0.6
+                )
+                ai_response = response[0]['generated_text'].replace(prompt, "").strip()
         else:
-            equity_pct, debt_pct = 60, 40
-            expected_return = "10-12%"
-            
-        equity_amount = investable_amount * (equity_pct / 100)
-        debt_amount = investable_amount * (debt_pct / 100)
-        
-        plan.append(f"Monthly Investment: ₹{investable_amount:,.0f}")
-        plan.append(f"Equity allocation: ₹{equity_amount:,.0f} ({equity_pct}%) - Nifty 50, Large Cap funds")
-        plan.append(f"Debt allocation: ₹{debt_amount:,.0f} ({debt_pct}%) - Debt funds, Bonds")
-        plan.append(f"Expected annual returns: {expected_return}")
-    
-    # Goal-based Planning
-    if "retirement" in profile.goals.lower():
-        plan.append(f"\n🏖️ **Retirement Planning**")
-        retirement_corpus = profile.monthly_income * 12 * 25
-        years_to_retire = 65 - profile.age
-        monthly_retirement_sip = retirement_corpus / (years_to_retire * 12) if years_to_retire > 0 else 0
-        plan.append(f"Target corpus: ₹{retirement_corpus:,.0f}")
-        plan.append(f"Monthly SIP needed: ₹{monthly_retirement_sip:,.0f}")
-        plan.append(f"Time horizon: {years_to_retire} years")
-    
-    if any(word in profile.goals.lower() for word in ["house", "home", "property"]):
-        plan.append(f"\n🏠 **Home Purchase Plan**")
-        home_value = profile.monthly_income * 12 * 5
-        down_payment = home_value * 0.2
-        years_to_buy = 7
-        monthly_home_save = down_payment / (years_to_buy * 12)
-        plan.append(f"Target home value: ₹{home_value:,.0f}")
-        plan.append(f"Down payment needed: ₹{down_payment:,.0f}")
-        plan.append(f"Monthly savings: ₹{monthly_home_save:,.0f} for {years_to_buy} years")
-    
-    # Insurance Strategy
-    if profile.user_type.lower() == "professional":
-        plan.append(f"\n🛡️ **Insurance Protection**")
-        term_coverage = profile.monthly_income * 12 * 10
-        term_premium_estimate = term_coverage / 1000  # Rough estimate
-        plan.append(f"Term life insurance: ₹{term_coverage:,.0f}")
-        plan.append(f"Estimated premium: ₹{term_premium_estimate:,.0f}/month")
-        plan.append(f"Health insurance: ₹10L family floater recommended")
-    
-    # Performance Analysis
-    plan.append(f"\n📈 **Current Performance Analysis**")
-    plan.append(f"Savings rate: {savings_rate:.1f}% {'✅ Excellent' if savings_rate > 30 else '⚠️ Needs improvement' if savings_rate > 20 else '🚨 Critical'}")
-    plan.append(f"Monthly surplus: ₹{savings:,.0f}")
-    
-    if savings_rate < 20:
-        plan.append(f"💡 **Immediate Action**: Increase savings rate to 25% minimum")
-        plan.append(f"Target monthly savings: ₹{profile.monthly_income * 0.25:,.0f}")
-    
-    return "\n".join(plan)
-
-def build_enhanced_prompt(profile: UserProfile, summary: Dict, user_query: str) -> str:
-    """Build enhanced prompt for AI with comprehensive context"""
-    
-    context = f"""
-FINANCIAL PROFILE:
-- User Type: {profile.user_type}
-- Monthly Income: ₹{profile.monthly_income:,.0f}
-- Risk Tolerance: {profile.risk_tolerance}
-- Financial Goals: {profile.goals}
-- Current Savings Rate: {summary['savings_rate_pct']:.1f}%
-- Monthly Surplus: ₹{summary['net_savings']:,.0f}
+            ai_response = "⚠️ No model available. Please check setup."
 
-EXPENSE ANALYSIS:
-- Total Monthly Expenses: ₹{summary['expense_total']:,.0f}
-- Top Spending Categories: {', '.join([f"{k}: ₹{v:,.0f}" for k, v in list(summary.get('top_expenses', {}).items())[:3]])}
+        st.session_state.chat_history.append({"role": "user", "content": user_input})
+        st.session_state.chat_history.append({"role": "ai", "content": ai_response})
 
-USER QUERY: {user_query}
-
-INSTRUCTIONS:
-As FinanceAI, provide specific, actionable financial advice for Indian markets. Include:
-1. Exact amounts and calculations
-2. Specific investment products (fund names, tax sections)
-3. Expected returns and time horizons
-4. Implementation steps
-5. Risk considerations
-
-Use current Indian financial regulations, tax laws (FY 2024-25), and market conditions.
-Be conversational but professional. Include emojis for better readability.
-"""
-    
-    return context
-
-INTENT_PATTERNS = {
-    "savings": r"save|savings|emergency|fd|rd|goal|corpus",
-    "tax": r"tax|80c|deduction|income tax|regime|tds|refund|section",
-    "invest": r"invest|sip|mutual fund|stock|index|portfolio|asset|equity|debt",
-    "budget": r"budget|spend|expense|track|summary|report|category",
-    "insurance": r"insurance|term|health|life|cover|premium|policy",
-    "retirement": r"retirement|pension|retire|old age|corpus",
-    "loan": r"loan|emi|mortgage|credit|debt|interest rate"
-}
-
-def detect_intent(text: str) -> str:
-    t = text.lower()
-    for intent, pattern in INTENT_PATTERNS.items():
-        if re.search(pattern, t):
-            return intent
-    return "general"
-
-# Modern Styling
-st.markdown("""
-<style>
-.main-header {
-    text-align: center;
-    padding: 2rem 0;
-    background: linear-gradient(90deg, #667eea 0%, #764ba2 100%);
-    -webkit-background-clip: text;
-    -webkit-text-fill-color: transparent;
-    font-size: 3.5rem;
-    font-weight: 800;
-    margin-bottom: 1rem;
-    text-shadow: 2px 2px 4px rgba(0,0,0,0.1);
-}
-.subtitle {
-    text-align: center;
-    color: #6b7280;
-    font-size: 1.3rem;
-    margin-bottom: 2rem;
-    font-weight: 500;
-}
-.ai-badge {
-    background: linear-gradient(45deg, #10b981, #3b82f6);
-    color: white;
-    padding: 0.5rem 1rem;
-    border-radius: 20px;
-    font-size: 0.9rem;
-    font-weight: 600;
-    display: inline-block;
-    margin: 0.5rem 0;
-}
-.financial-plan {
-    background: linear-gradient(135deg, #f0f9ff 0%, #e0f2fe 100%);
-    border-left: 5px solid #0ea5e9;
-    padding: 1.5rem;
-    margin: 1rem 0;
-    border-radius: 10px;
-    box-shadow: 0 4px 6px rgba(0,0,0,0.1);
-}
-.metric-card {
-    background: white;
-    border-radius: 12px;
-    padding: 1.5rem;
-    box-shadow: 0 2px 8px rgba(0,0,0,0.1);
-    border: 1px solid #e2e8f0;
-    text-align: center;
-}
-.chat-container {
-    max-width: 900px;
-    margin: 0 auto;
-    padding: 0 1rem;
-}
-</style>
-""", unsafe_allow_html=True)
-
-# Header
-st.markdown('<h1 class="main-header">🤖 FinanceAI</h1>', unsafe_allow_html=True)
-st.markdown('<p class="subtitle">AI-Powered Financial Advisor with GPT Intelligence</p>', unsafe_allow_html=True)
-
-# AI Status Badge
-if OPENAI_AVAILABLE and os.environ.get("OPENAI_API_KEY"):
-    st.markdown('<div class="ai-badge">🚀 Powered by OpenAI GPT-4 • Maximum Accuracy</div>', unsafe_allow_html=True)
-else:
-    st.markdown('<div class="ai-badge">⚡ Enhanced AI Fallback • Smart Responses</div>', unsafe_allow_html=True)
-
-# Quick Setup
-with st.expander("⚙️ Financial Profile Setup", expanded=True):
-    col1, col2, col3 = st.columns(3)
-    with col1:
-        monthly_income = st.number_input("💰 Monthly Income (₹)", min_value=0, value=50000, step=5000, help="Your total monthly earnings")
-        user_type = st.selectbox("👤 Profile Type", ["Professional", "Student"], index=0)
-    with col2:
-        risk = st.selectbox("📊 Risk Tolerance", ["Conservative", "Moderate", "Aggressive"], index=1, help="Your comfort level with investment risk")
-        uploaded = st.file_uploader("📄 Upload Transactions (CSV)", type=["csv"], help="Optional: Upload your expense data")
-    with col3:
-        provider_choice = st.selectbox("🤖 AI Engine", ["OpenAI GPT-4 (Recommended)", "Smart Fallback"], index=0)
-        goals = st.text_input("🎯 Financial Goals", value="Tax saving, wealth building, retirement planning", help="Your financial objectives")
-
-# Create profile
-profile = UserProfile(
-    name="User", user_type=user_type, age=28, country="India",
-    monthly_income=float(monthly_income), risk_tolerance=risk.lower(), goals=goals
-)
-
-# Initialize AI Providers
-if not st.session_state.provider_inited:
-    openai_provider = OpenAIProvider()
-    hf_provider = HuggingFaceProvider()
-    
-    if provider_choice.startswith("OpenAI") and openai_provider.available:
-        chosen_provider = openai_provider
-    else:
-        chosen_provider = hf_provider
-    
-    st.session_state.provider = chosen_provider
-    st.session_state.provider_inited = True
-
-provider = st.session_state.provider
-
-# Load and analyze financial data
-df = load_transactions(uploaded)
-summary = budget_summary(df, monthly_income_hint=profile.monthly_income)
-
-# Financial Overview
-st.markdown("### 📊 Financial Dashboard")
-col1, col2, col3, col4 = st.columns(4)
-
-with col1:
-    st.metric("💰 Monthly Income", f"₹{summary['income_total']:,.0f}", help="Total monthly earnings")
-with col2:
-    st.metric("💸 Expenses", f"₹{summary['expense_total']:,.0f}", help="Total monthly spending")
-with col3:
-    st.metric("🎯 Net Savings", f"₹{summary['net_savings']:,.0f}", help="Monthly surplus amount")
-with col4:
-    savings_color = "normal" if summary['savings_rate_pct'] > 20 else "inverse"
-    st.metric("📈 Savings Rate", f"{summary['savings_rate_pct']:.1f}%", help="Percentage of income saved")
-
-st.divider()
-
-# AI-Generated Financial Plan
-ai_plan = generate_ai_financial_plan(profile, summary)
-st.markdown("### 🎯 AI-Generated Financial Plan")
-st.markdown(f'<div class="financial-plan">{ai_plan}</div>', unsafe_allow_html=True)
-
-st.divider()
-
-# Enhanced Chat Interface
-st.markdown('<div class="chat-container">', unsafe_allow_html=True)
-st.markdown("### 💬 Ask Your Financial Questions")
-
-# Display chat history
-for turn in st.session_state.chat_history:
-    avatar = "🤖" if turn["role"] == "assistant" else "👤"
-    with st.chat_message(turn["role"], avatar=avatar):
-        st.markdown(turn["content"])
-
-# Chat input
-user_msg = st.chat_input("💬 Ask anything about investments, taxes, savings, or financial planning...")
-
-if user_msg:
-    # Add user message
-    st.session_state.chat_history.append({"role": "user", "content": user_msg})
-    
-    # Detect intent and build enhanced prompt
-    intent = detect_intent(user_msg)
-    enhanced_prompt = build_enhanced_prompt(profile, summary, user_msg)
-    
-    # Generate AI response
-    with st.chat_message("assistant", avatar="🤖"):
-        with st.spinner(f"🧠 {provider.name} is analyzing your query..."):
-            try:
-                ai_response = provider.generate(enhanced_prompt, max_tokens=1000)
-                
-                # Enhance response with intent-specific formatting
-                if intent == "tax":
-                    ai_response = f"📊 **Tax Optimization Advice**\n\n{ai_response}"
-                elif intent == "invest":
-                    ai_response = f"🚀 **Investment Strategy**\n\n{ai_response}"
-                elif intent == "savings":
-                    ai_response = f"💰 **Savings Plan**\n\n{ai_response}"
-                elif intent == "insurance":
-                    ai_response = f"🛡️ **Insurance Strategy**\n\n{ai_response}"
-                
-            except Exception as e:
-                ai_response = f"⚠️ **AI Service Temporarily Unavailable**\n\nI apologize for the inconvenience. Here's a quick response based on your query:\n\n"
-                
-                # Fallback responses based on intent
-                if intent == "tax":
-                    ai_response += "For tax optimization:\n• Invest ₹12,500/month in ELSS for Section 80C\n• Consider NPS for additional ₹50,000 deduction\n• Use tax-saving FDs for conservative investors"
-                elif intent == "invest":
-                    ai_response += f"For investment strategy:\n• Monthly SIP: ₹{summary['net_savings']*.6:.0f} in equity funds\n• Emergency fund: ₹{summary['net_savings']*.3:.0f} in liquid funds\n• Debt allocation: ₹{summary['net_savings']*.1:.0f} in debt funds"
-                else:
-                    ai_response += "Please check your internet connection and try again. For immediate help, consider:\n• Building 6-month emergency fund\n• Starting systematic investment plans\n• Optimizing tax-saving investments"
-        
-        st.markdown(ai_response)
-    
-    # Add AI response to history
-    st.session_state.chat_history.append({"role": "assistant", "content": ai_response})
-
-st.markdown('</div>', unsafe_allow_html=True)
+    # Display chat history
+    for msg in st.session_state.chat_history:
+        if msg["role"] == "user":
+            st.markdown(f"**👤 You:** {msg['content']}")
+        else:
+            st.markdown(f"**🤖 FinanceAI:** {msg['content']}")
 
-# Enhanced Footer
-st.markdown("""
-<div style="text-align: center; padding: 3rem 2rem; background: linear-gradient(135deg, #f8fafc 0%, #e2e8f0 100%); border-radius: 15px; margin-top: 3rem; border: 1px solid #cbd5e1;">
-    <h3 style="color: #1e293b; margin-bottom: 1rem;">🤖 FinanceAI</h3>
-    <p style="color: #475569; font-size: 1.1rem; margin-bottom: 0.5rem;"><strong>AI-Powered Financial Intelligence</strong></p>
-    <p style="color: #64748b; font-size: 0.95rem;">Powered by advanced AI models • Personalized recommendations • Always verify with current regulations</p>
-    <div style="margin-top: 1rem; font-size: 0.85rem; color: #94a3b8;">
-        🔒 Secure • 🎯 Accurate • 📊 Data-Driven • 🇮🇳 India-Focused
-    </div>
-</div>
-""", unsafe_allow_html=True)
+if __name__ == "__main__":
+    main()