Feat: Switch to Streamlit as main app (Option A)

Dockerfile

@@ -7,8 +7,8 @@ RUN pip install --no-cache-dir -r requirements.txt
 
 COPY . .
 
-# Expose API Port
+# Expose Streamlit Port
 EXPOSE 7860
 
-# Run FastAPI
-CMD ["uvicorn", "api.app:app", "--host", "0.0.0.0", "--port", "7860"]
+# Run Streamlit
+CMD ["streamlit", "run", "streamlit_app.py", "--server.port=7860", "--server.address=0.0.0.0", "--server.headless=true"]

streamlit_app.py

@@ -1,12 +1,11 @@
 """
-QuantScale AI - Streamlit Frontend
-Calls the existing FastAPI `/optimize` endpoint and displays the
-full portfolio allocation with Investment ($) and Allocation (%).
+QuantScale AI - Streamlit Frontend (Main App)
+Directly imports QuantScaleSystem - no HTTP dependency needed.
 """
 import re
-import requests
 import pandas as pd
 import streamlit as st
+from core.schema import OptimizationRequest
 
 # --- Page Config ---
 st.set_page_config(
@@ -16,20 +15,18 @@ st.set_page_config(
     initial_sidebar_state="collapsed"
 )
 
-# Custom dark-mode CSS
 st.markdown("""
 <style>
-    /* Overall dark background */
     .stApp { background-color: #0f1117; }
-    
-    /* Header */
     .main-header {
         background: linear-gradient(90deg, #60a5fa, #34d399);
         -webkit-background-clip: text;
+        background-clip: text;
         -webkit-text-fill-color: transparent;
         font-size: 2.5rem;
         font-weight: 700;
         text-align: center;
+        padding-top: 1rem;
     }
     .sub-header {
         color: #94a3b8;
@@ -37,26 +34,21 @@ st.markdown("""
         text-align: center;
         margin-bottom: 2rem;
     }
-    
-    /* Metric Cards */
     div[data-testid="metric-container"] {
         background-color: #1e212b;
         border: 1px solid #2d3748;
         border-radius: 12px;
         padding: 1rem;
     }
-    
-    /* Section headers */
     .section-title {
         color: #94a3b8;
-        font-size: 0.85rem;
+        font-size: 0.8rem;
         text-transform: uppercase;
         letter-spacing: 0.08em;
         font-weight: 600;
         margin-bottom: 0.5rem;
+        margin-top: 1.5rem;
     }
-    
-    /* Narrative / Commentary Box */
     .narrative-box {
         background-color: #1e212b;
         border-left: 4px solid #10b981;
@@ -64,21 +56,12 @@ st.markdown("""
         border-radius: 0 12px 12px 0;
         line-height: 1.8;
         color: #e2e8f0;
-    }
-
-    /* Dataframe styling override */
-    .stDataFrame thead tr th {
-        background-color: #1e212b !important;
-        color: #94a3b8 !important;
-        font-size: 0.8rem !important;
-        text-transform: uppercase;
-        letter-spacing: 0.05em;
+        font-size: 0.95rem;
     }
 </style>
 """, unsafe_allow_html=True)
 
 # --- Constants ---
-API_BASE_URL = "http://localhost:8000"  # Change to HF Space URL in prod
 SECTOR_KEYWORDS = {
     "Energy": ["energy", "oil", "gas"],
     "Technology": ["technology", "tech", "software", "it"],
@@ -93,32 +76,27 @@ SECTOR_KEYWORDS = {
 INCLUDE_KEYWORDS = ["keep", "include", "with", "stay", "portfolio", "only"]
 
 
+# --- Parsers ---
 def parse_investment_amount(text: str) -> float:
-    """Extract dollar amount from natural language. Returns 100_000 as default."""
-    text = text.replace(",", "")  # Remove commas: $10,000 -> $10000
-    # Match patterns like $10000, $10K, 10K, 10000, 50k
+    text = text.replace(",", "")
     match = re.search(r'\$?([\d.]+)\s*([kKmM]?)', text)
     if match:
         amount = float(match.group(1))
         suffix = match.group(2).lower()
-        if suffix == 'k':
-            amount *= 1_000
-        elif suffix == 'm':
-            amount *= 1_000_000
+        if suffix == 'k': amount *= 1_000
+        elif suffix == 'm': amount *= 1_000_000
         return amount
-    return 100_000.0  # Default
+    return 100_000.0
 
 
 def parse_excluded_sectors(text: str) -> list:
-    """Extract sectors to exclude from natural language, respecting 'keep' intent."""
     lower = text.lower()
     excluded = []
     for sector, keywords in SECTOR_KEYWORDS.items():
         if any(k in lower for k in keywords):
-            import re as _re
-            inc_pattern = _re.compile(
+            inc_pattern = re.compile(
                 rf'({"|".join(INCLUDE_KEYWORDS)})\s+(the\s+)?({"|".join([sector.lower()] + keywords)})',
-                _re.IGNORECASE
+                re.IGNORECASE
             )
             if not inc_pattern.search(lower):
                 excluded.append(sector)
@@ -126,10 +104,8 @@ def parse_excluded_sectors(text: str) -> list:
 
 
 def parse_strategy(text: str):
-    """Detect strategy keywords and Top N."""
     lower = text.lower()
-    strategy = None
-    top_n = None
+    strategy, top_n = None, None
     if "smallest" in lower:
         strategy = "smallest_market_cap"
     elif "largest" in lower:
@@ -141,25 +117,27 @@ def parse_strategy(text: str):
 
 
 def build_portfolio_df(allocations: dict, investment: float) -> pd.DataFrame:
-    """Convert raw allocation dict to a formatted DataFrame."""
     rows = []
     for ticker, weight in sorted(allocations.items(), key=lambda x: x[1], reverse=True):
         rows.append({
             "Ticker": ticker,
-            "Allocation (%)": weight,
-            "Investment ($)": weight * investment
+            "Allocation (%)": f"{weight * 100:.2f}%",
+            "Investment ($)": f"${weight * investment:,.2f}"
         })
-    df = pd.DataFrame(rows)
-    df["Allocation (%)"] = df["Allocation (%)"].apply(lambda x: f"{x * 100:.2f}%")
-    df["Investment ($)"] = df["Investment ($)"].apply(lambda x: f"${x:,.2f}")
-    return df
+    return pd.DataFrame(rows)
+
 
+# --- Lazy-load system to avoid import overhead on every rerender ---
+@st.cache_resource(show_spinner="Loading QuantScale Engine...")
+def get_system():
+    from main import QuantScaleSystem
+    return QuantScaleSystem()
 
-# --- UI Layout ---
+
+# --- UI ---
 st.markdown('<div class="main-header">QuantScale AI</div>', unsafe_allow_html=True)
 st.markdown('<div class="sub-header">Direct Indexing & Attribution Engine</div>', unsafe_allow_html=True)
 
-# Input
 user_input = st.text_area(
     "",
     placeholder="Describe your goal, e.g., 'Optimize my $10,000 portfolio but exclude the Energy sector.'",
@@ -168,85 +146,84 @@ user_input = st.text_area(
 )
 run_btn = st.button("🚀 Generate Portfolio Strategy", use_container_width=True, type="primary")
 
-# --- Main Logic ---
 if run_btn and user_input:
     investment_amount = parse_investment_amount(user_input)
     excluded_sectors = parse_excluded_sectors(user_input)
     strategy, top_n = parse_strategy(user_input)
 
-    payload = {
-        "client_id": "StreamlitUser",
-        "initial_investment": investment_amount,
-        "excluded_sectors": excluded_sectors,
-        "excluded_tickers": [],
-        "strategy": strategy,
-        "top_n": top_n,
-        "benchmark": "^GSPC"
-    }
-
-    with st.spinner("Running Convex Optimization & AI Analysis..."):
+    request = OptimizationRequest(
+        client_id="StreamlitUser",
+        initial_investment=investment_amount,
+        excluded_sectors=excluded_sectors,
+        excluded_tickers=[],
+        strategy=strategy,
+        top_n=top_n,
+        benchmark="^GSPC"
+    )
+
+    with st.spinner("⚙️ Running Convex Optimization & AI Analysis..."):
         try:
-            resp = requests.post(f"{API_BASE_URL}/optimize", json=payload, timeout=120)
-            resp.raise_for_status()
-            data = resp.json()
-        except requests.exceptions.RequestException as e:
-            st.error(f"❌ API Error: {e}")
+            system = get_system()
+            result = system.run_pipeline(request)
+        except Exception as e:
+            st.error(f"❌ Optimization error: {e}")
             st.stop()
 
-    # --- Metrics Row ---
+    if not result:
+        st.error("Pipeline returned no result. Check your input.")
+        st.stop()
+
+    opt = result["optimization"]
+    commentary = result["commentary"]
+
+    # --- Metrics ---
     col1, col2, col3 = st.columns(3)
     with col1:
-        st.metric(
-            "💼 Invested Amount",
-            f"${investment_amount:,.0f}"
-        )
+        st.metric("💼 Invested", f"${investment_amount:,.0f}")
     with col2:
-        te = data.get("tracking_error", 0.0)
         st.metric(
-            "📊 Projected Tracking Error",
-            f"{te * 100:.4f}%",
-            help="How closely this portfolio tracks the S&P 500. Lower = more index-like."
+            "📊 Tracking Error",
+            f"{opt.tracking_error * 100:.4f}%",
+            help="How closely the portfolio tracks the S&P 500"
         )
     with col3:
-        excl_text = ", ".join(excluded_sectors) if excluded_sectors else "None"
-        st.metric("🚫 Excluded Sectors", excl_text if len(excl_text) < 25 else f"{len(excluded_sectors)} Sectors")
+        excl_display = ", ".join(excluded_sectors) if excluded_sectors else "None"
+        st.metric("🚫 Excluded", excl_display if len(excl_display) <= 30 else f"{len(excluded_sectors)} Sectors")
 
     st.divider()
 
     # --- AI Commentary ---
     st.markdown('<p class="section-title">AI Performance Attribution</p>', unsafe_allow_html=True)
-    narrative = data.get("attribution_narrative", "No commentary generated.")
-    st.markdown(f'<div class="narrative-box">{narrative}</div>', unsafe_allow_html=True)
+    st.markdown(f'<div class="narrative-box">{commentary}</div>', unsafe_allow_html=True)
 
     st.divider()
 
     # --- Full Portfolio Table ---
-    allocations = data.get("allocations", {})
+    allocations = opt.weights
     if allocations:
         df = build_portfolio_df(allocations, investment_amount)
-        total_holdings = len(df)
-        
+        total = len(df)
+
         st.markdown(
-            f'<p class="section-title">Full Portfolio Allocation (100%) — {total_holdings} Holdings</p>',
+            f'<p class="section-title">Full Portfolio Allocation (100%) — {total} Holdings</p>',
             unsafe_allow_html=True
         )
-        
-        # Summary stats row above table
+
         c1, c2, c3 = st.columns(3)
-        c1.metric("Total Holdings", total_holdings)
-        c2.metric("Largest Position", df["Ticker"].iloc[0] if len(df) else "—")
-        c3.metric("Smallest Position", df["Ticker"].iloc[-1] if len(df) else "—")
-        
+        c1.metric("Total Holdings", total)
+        c2.metric("Largest Position", df["Ticker"].iloc[0])
+        c3.metric("Smallest Position", df["Ticker"].iloc[-1])
+
         st.dataframe(
             df,
             use_container_width=True,
             hide_index=True,
-            height=min(400, 36 * total_holdings + 36),  # Dynamic auto-height, max 400px scrollable
+            height=min(500, 36 * total + 40),
             column_config={
                 "Ticker": st.column_config.TextColumn("Ticker", width="small"),
-                "Allocation (%)": st.column_config.TextColumn("Allocation (%)", width="medium"),
-                "Investment ($)": st.column_config.TextColumn(f"Investment ($) of ${investment_amount:,.0f}", width="medium"),
+                "Allocation (%)": st.column_config.TextColumn("Allocation (%)", width="small"),
+                "Investment ($)": st.column_config.TextColumn(
+                    f"Investment (of ${investment_amount:,.0f})", width="medium"
+                ),
             }
         )
-    else:
-        st.warning("No allocation data returned from the optimizer.")