Update app.py

app.py

@@ -1681,6 +1681,312 @@ async def health_check():
     return {"status": "healthy", "service": "Real Estate Financial Model API"}
     
 
+@app.post("/api/generate-model-with-summary")
+async def generate_model_with_summary(files: List[UploadFile] = File(...)):
+    """
+    API endpoint that returns JSON summary with key metrics
+    AND includes base64 encoded Excel file
+    
+    Returns JSON with:
+    - Key financial metrics
+    - Deal summary
+    - Market analysis
+    - Excel file as base64
+    """
+    if not files:
+        raise HTTPException(status_code=400, detail="No files uploaded")
+    
+    temp_dir = None
+    try:
+        # Create temporary directory
+        temp_dir = tempfile.mkdtemp()
+        print(f"📁 Created temp directory: {temp_dir}")
+        
+        # Save uploaded files
+        for upload_file in files:
+            file_path = Path(temp_dir) / upload_file.filename
+            print(f"💾 Saving file: {upload_file.filename}")
+            
+            with open(file_path, "wb") as f:
+                content = await upload_file.read()
+                f.write(content)
+        
+        print(f"✅ Saved {len(files)} files")
+        
+        # Initialize pipeline
+        print("🔧 Initializing pipeline...")
+        pipeline = RealEstateModelPipeline(GEMINI_API_KEY)
+        
+        # Create output file
+        output_file = Path(temp_dir) / "Real_Estate_Financial_Model.xlsx"
+        
+        # Run pipeline
+        print("🚀 Running pipeline...")
+        pipeline.run_full_pipeline(temp_dir, str(output_file))
+        
+        # Check if file was created
+        if not output_file.exists():
+            raise Exception("Excel file was not generated")
+        
+        print(f"✅ Excel file created: {output_file}")
+        
+        # Get formula results
+        r = pipeline.formula_results
+        d = pipeline.structured_data
+        
+        # Build comprehensive JSON response
+        response_data = {
+            "status": "success",
+            "message": "Financial model generated successfully",
+            
+            # KEY METRICS (Dashboard Top Cards)
+            "key_metrics": {
+                "purchase_price": {
+                    "value": r.get('PRICE', 0),
+                    "formatted": f"${r.get('PRICE', 0):,.0f}",
+                    "label": "Purchase Price"
+                },
+                "noi": {
+                    "value": r.get('NET_OPERATING_INCOME', 0),
+                    "formatted": f"${r.get('NET_OPERATING_INCOME', 0):,.0f}",
+                    "label": "NOI"
+                },
+                "cap_rate": {
+                    "value": r.get('CAP_RATE', 0),
+                    "formatted": f"{r.get('CAP_RATE', 0):.1f}%",
+                    "label": "Cap Rate"
+                },
+                "irr_10yr": {
+                    "value": float(r.get('IRR_TO_LP', 0).real if isinstance(r.get('IRR_TO_LP', 0), complex) else r.get('IRR_TO_LP', 0)),
+                    "formatted": f"{float(r.get('IRR_TO_LP', 0).real if isinstance(r.get('IRR_TO_LP', 0), complex) else r.get('IRR_TO_LP', 0)):.1f}%",
+                    "label": "IRR (10 yr)"
+                },
+                "dscr": {
+                    "value": r.get('DEBT_SERVICE_COVERAGE_RATIO', 0),
+                    "formatted": f"{r.get('DEBT_SERVICE_COVERAGE_RATIO', 0):.2f}x",
+                    "label": "DSCR"
+                }
+            },
+            
+            # T12 SUMMARY (Extracted Summary)
+            "t12_summary": {
+                "gross_income": {
+                    "annual": r.get('EFFECTIVE_GROSS_INCOME', 0),
+                    "monthly": r.get('EFFECTIVE_GROSS_INCOME', 0) / 12,
+                    "source": "Calculated from rent roll"
+                },
+                "operating_expenses": {
+                    "annual": r.get('TOTAL_OPERATING_EXPENSES', 0),
+                    "monthly": r.get('TOTAL_OPERATING_EXPENSES', 0) / 12,
+                    "expense_ratio": f"{(r.get('TOTAL_OPERATING_EXPENSES', 0) / r.get('EFFECTIVE_GROSS_INCOME', 1) if r.get('EFFECTIVE_GROSS_INCOME', 0) > 0 else 0):.1%}"
+                },
+                "noi": {
+                    "annual": r.get('NET_OPERATING_INCOME', 0),
+                    "monthly": r.get('NET_OPERATING_INCOME', 0) / 12,
+                    "verified": True
+                },
+                "occupancy": {
+                    "rate": f"{(1 - r.get('VACANCY_RATE', 0)):.0%}",
+                    "value": 1 - r.get('VACANCY_RATE', 0)
+                }
+            },
+            
+            # PROPERTY INFO
+            "property_info": {
+                "address": d.get('property_info', {}).get('address', 'N/A'),
+                "property_type": d.get('property_info', {}).get('property_type', 'N/A'),
+                "units": r.get('UNITS', 0),
+                "gross_sf": r.get('GROSS_SF', 0),
+                "rentable_sf": r.get('RENTABLE_SF', 0),
+                "building_efficiency": f"{r.get('BUILDING_EFFICIENCY', 0):.1%}"
+            },
+            
+            # DEAL VS MARKET AVERAGES
+            "market_comparison": {
+                "your_deal": {
+                    "cap_rate": r.get('CAP_RATE', 0) / 100,
+                    "irr": float(r.get('IRR_TO_LP', 0).real if isinstance(r.get('IRR_TO_LP', 0), complex) else r.get('IRR_TO_LP', 0)) / 100,
+                    "dscr": r.get('DEBT_SERVICE_COVERAGE_RATIO', 0)
+                },
+                "market_average": {
+                    "cap_rate": 0.055,  # Example market average
+                    "irr": 0.12,
+                    "dscr": 1.25
+                }
+            },
+            
+            # AI OBSERVATIONS
+            "ai_observations": [
+                {
+                    "observation": f"Cap Rate {'above' if r.get('CAP_RATE', 0) > 5.5 else 'below'} area average",
+                    "recommendation": "undervalued potential" if r.get('CAP_RATE', 0) > 5.5 else "premium pricing"
+                },
+                {
+                    "observation": f"Expense ratio at {(r.get('TOTAL_OPERATING_EXPENSES', 0) / r.get('EFFECTIVE_GROSS_INCOME', 1) if r.get('EFFECTIVE_GROSS_INCOME', 0) > 0 else 0):.1%}",
+                    "recommendation": "within normal range" if 0.3 <= (r.get('TOTAL_OPERATING_EXPENSES', 0) / r.get('EFFECTIVE_GROSS_INCOME', 1)) <= 0.45 else "review expenses"
+                },
+                {
+                    "observation": "DSCR meets financing standards" if r.get('DEBT_SERVICE_COVERAGE_RATIO', 0) >= 1.2 else "DSCR below standards",
+                    "recommendation": "approved" if r.get('DEBT_SERVICE_COVERAGE_RATIO', 0) >= 1.2 else "increase equity"
+                },
+                {
+                    "observation": "NOI verification",
+                    "recommendation": "Full underwriting review recommended"
+                }
+            ],
+            
+            # CAPITAL STACK
+            "capital_stack": {
+                "total_project_cost": r.get('TOTAL_FINANCING_CONTINGENCY_AND_RESERVES', 0),
+                "total_debt": r.get('TOTAL_DEBT', 0),
+                "total_equity": r.get('TOTAL_EQUITY', 0),
+                "debt_percentage": r.get('DEBT_PERCENTAGE', 0),
+                "equity_percentage": r.get('EQUITY_PERCENTAGE', 0),
+                "ltc_ratio": r.get('LTC_RATIO', 0)
+            },
+            
+            # RETURNS ANALYSIS
+            "returns": {
+                "yield_on_cost": r.get('YIELD_ON_COST_PERCENTAGE', 0),
+                "stabilized_yield_on_cost": r.get('STABILIZED_YIELD_ON_COST_PERCENTAGE', 0),
+                "cash_on_cash_return": r.get('CASH_ON_CASH_RETURN', 0),
+                "lp_multiple": r.get('LP_MULTIPLE', 0),
+                "lp_irr": float(r.get('IRR_TO_LP', 0).real if isinstance(r.get('IRR_TO_LP', 0), complex) else r.get('IRR_TO_LP', 0)),
+                "hold_period_years": r.get('HOLD_PERIOD_MONTHS', 0) / 12
+            },
+            
+            # OPERATING METRICS
+            "operating_metrics": {
+                "gross_potential_rent": r.get('GROSS_POTENTIAL_FREE_MARKET_RENT', 0) + r.get('GROSS_POTENTIAL_AFFORDABLE_RENT', 0),
+                "vacancy_loss": r.get('VACANCY_LOSS', 0),
+                "effective_gross_income": r.get('EFFECTIVE_GROSS_INCOME', 0),
+                "operating_expenses": r.get('TOTAL_OPERATING_EXPENSES', 0),
+                "noi": r.get('NET_OPERATING_INCOME', 0),
+                "noi_per_unit": r.get('NOI_PER_UNIT', 0),
+                "noi_per_sf": r.get('NOI_PER_GSF', 0)
+            },
+            
+            # COST BREAKDOWN
+            "cost_breakdown": {
+                "acquisition": {
+                    "total": r.get('TOTAL_ACQUISITION_COST', 0),
+                    "per_unit": r.get('TOTAL_ACQUISITION_COST_PER_UNIT', 0),
+                    "per_sf": r.get('TOTAL_ACQUISITION_COST_PER_GSF', 0)
+                },
+                "construction": {
+                    "total": r.get('TOTAL_CONSTRUCTION_GMP', 0),
+                    "per_unit": r.get('CONSTRUCTION_GMP_PER_UNIT', 0),
+                    "per_sf": r.get('CONSTRUCTION_GMP_PER_GSF', 0)
+                },
+                "soft_costs": {
+                    "total": r.get('TOTAL_SOFT_COST', 0),
+                    "per_sf": r.get('TOTAL_SOFT_COST_PER_GSF', 0)
+                },
+                "contingency": r.get('CONTINGENCY_COST', 0),
+                "reserves": r.get('OPERATING_RESERVE', 0)
+            },
+            
+            # SOURCE FILES
+            "source_files": [
+                {
+                    "filename": file.filename,
+                    "size": len(await file.read()) if hasattr(file, 'read') else 0,
+                    "type": file.filename.split('.')[-1].upper()
+                }
+                for file in files
+            ],
+            
+            # EXCEL FILE (BASE64 ENCODED)
+            "excel_file": {
+                "filename": "Real_Estate_Financial_Model.xlsx",
+                "size_bytes": output_file.stat().st_size,
+                "base64_data": base64.b64encode(output_file.read_bytes()).decode('utf-8')
+            }
+        }
+        
+        print("✅ Response JSON created")
+        
+        return JSONResponse(content=response_data)
+        
+    except Exception as e:
+        import traceback
+        error_details = traceback.format_exc()
+        print("❌ ERROR OCCURRED:")
+        print(error_details)
+        
+        raise HTTPException(
+            status_code=500, 
+            detail={
+                "error": str(e),
+                "type": type(e).__name__,
+                "traceback": error_details
+            }
+        )
+
+
+
+# NEW: Endpoint that returns ONLY JSON (no Excel file)
+@app.post("/api/analyze-only")
+async def analyze_only(files: List[UploadFile] = File(...)):
+    """
+    Returns only JSON summary without Excel file
+    Faster response, good for dashboards
+    """
+    if not files:
+        raise HTTPException(status_code=400, detail="No files uploaded")
+    
+    temp_dir = None
+    try:
+        temp_dir = tempfile.mkdtemp()
+        
+        for upload_file in files:
+            file_path = Path(temp_dir) / upload_file.filename
+            with open(file_path, "wb") as f:
+                content = await upload_file.read()
+                f.write(content)
+        
+        pipeline = RealEstateModelPipeline(GEMINI_API_KEY)
+        
+        # Extract and process data (no Excel generation)
+        pipeline.extract_all_pdfs(temp_dir)
+        structured_data = pipeline.extract_structured_data()
+        structured_data = pipeline.post_process_extracted_data(structured_data)
+        pipeline.calculate_all_formulas(structured_data)
+        
+        r = pipeline.formula_results
+        d = pipeline.structured_data
+        
+        # Return lightweight JSON
+        return JSONResponse(content={
+            "status": "success",
+            "key_metrics": {
+                "purchase_price": r.get('PRICE', 0),
+                "noi": r.get('NET_OPERATING_INCOME', 0),
+                "cap_rate": r.get('CAP_RATE', 0),
+                "irr": float(r.get('IRR_TO_LP', 0).real if isinstance(r.get('IRR_TO_LP', 0), complex) else r.get('IRR_TO_LP', 0)),
+                "dscr": r.get('DEBT_SERVICE_COVERAGE_RATIO', 0)
+            },
+            "property_info": {
+                "address": d.get('property_info', {}).get('address', 'N/A'),
+                "units": r.get('UNITS', 0),
+                "gross_sf": r.get('GROSS_SF', 0)
+            },
+            "capital_stack": {
+                "total_project_cost": r.get('TOTAL_FINANCING_CONTINGENCY_AND_RESERVES', 0),
+                "total_debt": r.get('TOTAL_DEBT', 0),
+                "total_equity": r.get('TOTAL_EQUITY', 0)
+            },
+            "returns": {
+                "yield_on_cost": r.get('YIELD_ON_COST_PERCENTAGE', 0),
+                "cash_on_cash": r.get('CASH_ON_CASH_RETURN', 0),
+                "lp_multiple": r.get('LP_MULTIPLE', 0)
+            }
+        })
+        
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+
 if __name__ == "__main__":
     
     # Hardcoded API Key