RealtyAI - Smart Real Estate Insight Platform

A comprehensive AI-powered real estate analytics platform that provides price predictions and time series forecasting for various regions using Machine Learning and Prophet models.

Table of Contents

• Features
• Technology Stack
• Project Structure
• Prerequisites
• Installation & Setup
• Running the Application
• Models
• License

Features

Price Prediction

• Predict real estate prices based on property features
• Input parameters: Location, City, BHK, Total Area, Price per SQFT, Bathrooms, Balcony
• Uses BaggingRegressor ML pipeline for accurate predictions

Time Series Forecasting

• Single Region Forecast: Detailed forecast with confidence intervals for one region
• Multi-Region Comparison: Compare forecasts across multiple regions
• Region Statistics: Historical data analysis and market insights
• Forecast horizon: 1-36 months
• Prophet-based forecasting models

Visualization

• Interactive charts with Recharts
• Historical data vs. forecast comparison
• Confidence interval visualization
• Responsive design for all devices

Technology Stack

Backend

• Framework: FastAPI 0.104.0+
• ML Libraries:
  • scikit-learn 1.7.1
  • Prophet 1.1.0+
  • pandas 2.0.0+
  • numpy 1.24.0+
• Model Serialization: joblib 1.3.0+
• Server: Uvicorn (with standard extras)

Frontend

• Framework: React 18
• Charts: Recharts
• Icons: React Icons (Font Awesome)
• HTTP Client: Fetch API
• Build Tool: Create React App

Package Management

• Backend: uv (Astral's fast Python package installer)
• Frontend: npm/yarn

Project Structure

RealtyAI_Infosys_Internship_Aug2025/
│
├── backend/
│   ├── main.py                    # FastAPI application
│   ├── feature_engineering.py     # Feature engineering transformer
│   ├── pyproject.toml            # uv dependencies
│   └── README.md                 # Backend documentation
│
├── frontend/
│   ├── public/
│   │   ├── index.html
│   │   └── manifest.json
│   ├── src/
│   │   ├── components/
│   │   │   ├── PriceForecasting.js   # Time series forecasting UI
│   │   │   └── PricePrediction.js    # Price prediction UI
│   │   ├── App.js
│   │   ├── App.css
│   │   ├── index.js
│   │   └── index.css
│   ├── package.json
│   └── README.md
│
├── Models/
│   ├── real_estate_pipeline_v20250915_182141.joblib   # Price prediction model
│   └── all_region_models.joblib                        # Prophet time series models
│
├── Notebooks/
│   ├── RealEstate_Feature_Engineering_and_training.ipynb
│   ├── Time_Series_Fore_Casting.ipynb
│   ├── EDA_price_prediction.ipynb
│   └── ... (other analysis notebooks)
│
├── AI Project_ RealtyAI Smart Real Estate Insight Platform.pdf
├── LICENSE
└── README.md                      # This file

Prerequisites

Before you begin, ensure you have the following installed:

System Requirements

• Operating System: Windows 10/11, macOS, or Linux
• Python: 3.10 or higher (required for scikit-learn 1.7.1)
• Node.js: 16.x or higher
• npm: 8.x or higher (comes with Node.js)

Package Managers

• uv: Fast Python package installer (Installation Guide)
• npm/yarn: For frontend dependencies

Installation & Setup

Step 1: Clone the Repository

cd d:\dev\test\internship
git clone https://github.com/AabidMK/RealtyAI_Infosys_Internship_Aug2025.git
cd RealtyAI_Infosys_Internship_Aug2025

Step 2: Backend Setup

2.1 Install uv (Python Package Manager)

For Windows:

# Download and run the installer
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

For macOS/Linux:

curl -LsSf https://astral.sh/uv/install.sh | sh

Verify installation:

uv --version

2.2 Install Backend Dependencies

cd backend
uv sync

This will install all required packages from pyproject.toml:

• FastAPI
• Uvicorn with standard extras
• scikit-learn 1.7.1 (exact version for model compatibility)
• Prophet
• pandas, numpy, joblib, pydantic

2.3 Verify Model Files

Ensure the following model files exist in the Models/ directory:

Models/
├── real_estate_pipeline_v20250915_182141.joblib
└── all_region_models.joblib

If missing, download from the project repository or train new models using the provided notebooks.

Step 3: Frontend Setup

3.1 Navigate to Frontend Directory

cd ../frontend

3.2 Install Node Dependencies

npm install

This will install:

• React and React-DOM
• Recharts
• React Icons
• Other development dependencies

Running the Application

Start Backend Server

Open a terminal and run:

cd backend
uv run uvicorn main:app --reload --host 127.0.0.1 --port 8000

Expected Output:

INFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
INFO:     Started reloader process
INFO:     Started server process
INFO:     Waiting for application startup.
INFO:     Application startup complete.

Backend will be available at: http://127.0.0.1:8000

Start Frontend Development Server

Open a new terminal and run:

cd frontend
npm start

Expected Output:

Compiled successfully!

You can now view frontend in the browser.

  Local:            http://localhost:3000
  On Your Network:  http://192.168.x.x:3000

Frontend will be available at: http://localhost:3000

Models

Price Prediction Model

• Algorithm: BaggingRegressor with Decision Tree base estimators
• Features: Location, City, BHK, Total_Area, Price_per_SQFT, Bathroom, Balcony
• Training Data: Real estate listings from multiple Indian cities
• File: real_estate_pipeline_v20250915_182141.joblib

Time Series Forecasting Models

• Algorithm: Facebook Prophet
• Regions: 50+ US states/regions
• Training Period: 1996-2018 (historical ZHVI data)
• Forecast Capability: Up to 36 months ahead
• File: all_region_models.joblib

License

This project is licensed under the MIT License - see the LICENSE file for details.