🏠 LensEstate — AI Solution for Tunisian Real Estate Market Analysis

Overview

TuniEstate is an AI-powered platform designed to address the lack of transparency and reliable data in the Tunisian real estate market. By combining autonomous data ingestion agents, machine learning models, and a generative AI assistant, TuniEstate centralizes fragmented property listings, estimates fair market prices, and delivers actionable insights to buyers, sellers, investors, and agencies — all through an intuitive, accessible interface.

---

Features

• 🤖 AI Data Collection Agents — AI agents autonomously collect, clean, deduplicate, and normalize real estate data from multiple sources (online platforms, social media, informal listings)
• 💰 Multi-Modal Valuation Models — ML models combining property features, location, and market data to ensure fair and objective pricing
• 📈 Regional Trend Prediction — Time-series models to anticipate price shifts and market evolution across Tunisian regions
• 💬 Conversational GenAI Assistant — Natural language chatbot that simplifies market access and delivers personalized insights on demand
• 🏗️ 2D Property Reconstruction — Visual reconstruction of property layouts to improve how properties are presented and understood

---

Tech Stack

Frontend

Technology	Purpose
React + TypeScript	UI Framework
Vite	Build Tool
TanStack Router	Routing
Tailwind CSS	Styling
Framer Motion	Animations
Bun	Package Manager

Backend

Technology	Purpose
Django + DRF	REST API Framework
PostgreSQL	Database
ChromaDB	Vector Database (RAG)
Sentence Transformers	Embeddings
LLaMA 3.1 70B	AI Language Model
OpenAI SDK	LLM API Client
JWT	Authentication
ReportLab	PDF Generation

---

Architecture

LensEstate/
├── frontend/                  # Client-side application
│   ├── src/
│   │   ├── components/        # Reusable UI components
│   │   ├── routes/            # TanStack Router pages
│   │   ├── hooks/             # Custom React hooks
│   │   ├── context/           # Auth context
│   │   ├── lib/               # API client & utilities
│   │   └── types/             # TypeScript types
│   ├── vite.config.ts
│   └── package.json
├── backend/                   # Server-side application
│   ├── users/                 # Auth & AI chatbot
│   ├── listings/              # Real estate listings API
│   ├── predictions/           # ML price prediction
│   ├── forecasting/           # Market forecasting
│   ├── agents/                # Autonomous data agents
│   ├── plans/                 # Investment plans
│   └── lenstate/              # Django settings & URLs
├── ma_base_immo/              # ChromaDB vector database
├── pdf_generator.py           # PDF report generation
├── .env.local                 # Frontend environment variables
└── README.md

---

Contributors

Name	Role
Ben Dhifallah Ines	Project Manager
Mohamed Khalil Kamessi	Solution Architect
Mohamed Ali Jabri	Data Scientist
Najeh Ben Rebah	Project Lead
Feryel Ifaoui	Data Scientist

---

Academic Context

🎓 This project was developed as part of a Data Science academic program at Esprit School of Engineering, during the academic year 2025–2026.

Supervisor: Mme Wiem Trabelsi & Mr Bilel Farah

This project aligns with the following UN Sustainable Development Goals:

SDG	Relevance
🎓 SDG 4 — Quality Education	Transforms complex market data into accessible insights
💼 SDG 8 — Decent Work & Economic Growth	Supports informed real estate decision-making
🏭 SDG 9 — Industry, Innovation & Infrastructure	Applies AI/ML to structure Tunisian real estate data
🏙️ SDG 11 — Sustainable Cities & Communities	Supports urban development analysis across regions

---

Getting Started

Prerequisites

• Python >= 3.11
• Node.js >= 18
• Bun >= 1.3
• PostgreSQL >= 18

Installation

# Clone the repository
git clone https://github.com/InesBenDhifallah/-Esprit-PIDS-4DS6--2026-LensEstate.git
cd -Esprit-PIDS-4DS6--2026-LensEstate

# Create and activate virtual environment
python -m venv venv
venv\Scripts\Activate.ps1  # Windows

# Install backend dependencies
cd backend && pip install -r requirements.txt

# Install frontend dependencies
cd ../frontend && bun install

Running the App

# Start the backend
cd backend && python manage.py runserver

# Start the frontend (in a separate terminal)
cd frontend && bun run dev

Environment Variables

Create a .env file in the backend/ directory:

SECRET_KEY=your_secret_key
DEBUG=True
DB_NAME=LensEstate
DB_USER=your_db_user
DB_PASSWORD=your_db_password
DB_HOST=localhost
DB_PORT=5432
EMAIL_HOST_USER=your_email
EMAIL_HOST_PASSWORD=your_email_password
GOOGLE_CLIENT_ID=your_google_client_id
GOOGLE_CLIENT_SECRET=your_google_client_secret

Create a .env.local file in the root directory:

API_KEY=your_api_key
VITE_API_URL=http://localhost:8000
VITE_GOOGLE_CLIENT_ID=your_google_client_id

Acknowledgments

• Tunisian real estate data sources and online listing platforms
• Open-source ML and NLP libraries powering the analytical engine

---

Built with ❤️ by the LensEstate team — making the Tunisian real estate market transparent, fair, and accessible for all.