Apppublic

SagarKeshave/real-estate-atlas

sourceHugging Faceupdated 11mo agoView on Hugging Face

0likes

App README

Real Estate Atlas RAG Chatbot 🏡

Flask application for real-estate insights and conversational assistance, powered by LangChain, Gemini, and Chroma DB for retrieval-augmented generation.

🎥 Demo Video: Watch on Loom

About

Real Estate Atlas is a conversational assistant (chatbot) tailored for real estate domain queries. It uses Retrieval-Augmented Generation (RAG) to fetch domain-specific data (via Chroma DB) and combine with generative models (LangChain + Gemini) to deliver informed responses.

Features

—Conversational Q&A for real estate topics
—Retrieval of domain data (e.g. property features, market trends)
—Integration with generative model (Gemini) to produce fluent, context-aware replies
—Dockerized deployment for easy setup
—Modular architecture for ingestion, preprocessing, and model orchestration

Setup & Installation

1.Clone the repository:

bash

   git clone https://github.com/Sagarkeshave/Real-Estate-Atlas.git
   cd Real-Estate-Atlas

1.Create a virtual environment

bash

    python3 -m venv venv
    source venv/bin/activate

1.Install dependencies:

bash

    pip install -r requirements.txt

1.Ensure configuration / secrets are set # In .env

bash

    GOOGLE_API_KEY = "YOUR_API_KEY"

Usage

Once running, you can send queries via the frontend interface or via HTTP API endpoints (e.g. /chat or similar). The system will:

1.Ingest the user prompt

1.Use Chroma DB to find relevant document/context

1.Combine context + prompt and send to Gemini model

1.Return generated answer to user

Project Structure

bash

├── app.py
├── st_app.py
├── dataIngestion.py
├── dataPreprocessing.py
├── rag.py
├── requirements.txt
├── Dockerfile
├── templates/
├── static/
└── vectore_DB/ chroma_langchain_DB/