HomeGPT

Welcome to HomeGPT, an innovative project designed to help homeowners, buyers, and real estate professionals understand complex home inspection reports using cutting-edge technology. By leveraging Language Models (LLM) and Retrieval-Augmented Generation (RAG), HomeGPT interprets technical jargon and presents it in an easy-to-understand format.

Overview

Home inspection reports are crucial for making informed decisions in real estate transactions but often contain technical language that can be difficult to understand. HomeGPT aims to bridge this gap by providing clear, concise, and actionable insights from these reports.

Features

• Report Summarization: Automatically summarizes detailed reports into digestible insights.
• Jargon Explanation: (To Be Build) Explains technical terms and conditions in simple language.
• Recommendation System: (To Be Build) Suggests potential next steps based on the inspection findings.
• Interactive Queries: (To Be Build) Allows users to ask specific questions about the report for detailed answers.

Dataset

Complext home report data

Technology

open HomeGPT combines the power of Large Language Models (LLM) with Retrieval-Augmented Generation (RAG) to provide accurate and contextually relevant information:

• LLM: Processes natural language data to understand and generate human-like text.
• RAG: Enhances the model's responses by retrieving relevant information from a large database of home inspection terminologies and prior reports.

Pipeline structure

Content

1. CBHI Home Inspection Report.pdf contains a sample inspection pdf.
2. ExtractDatafromHomeReportPDF.ipynb contains code for generating the HomeGPT inspection report.
3. HomeGPT_vision_module.ipynb contains future potential involving image understanding.