🏘️ CompWise_AI

CompWise_AI is an AI-powered dashboard for analyzing and explaining comparable property rankings based on machine learning and natural language explanations. Forked and customized from an original ML property-ranking system, this version emphasizes interactivity and visualization with a polished Streamlit frontend.

---

🚀 Features

• 📊 Visualizes comparable property data
• 🧠 Uses GPT-3.5 to generate explanations for property ranking
• 📝 Collects user feedback on suggestions
• 💡 Interactive and intuitive UI using Streamlit

---

🧠 How It Works

1. Input Data: Loads top3_gpt_explanations.csv which includes candidate property data and GPT-generated explanations.
2. Filtering: Users can select specific appraisal orders, apply score thresholds, and filter by bedroom count.
3. Visualization:
   • Bar charts for score ranking and price distribution
   • Feature-level comparisons between subject and candidate properties
4. Feedback Loop: Users provide input on whether the ranked comparables are relevant, which gets logged to feedback_log.csv.
5. Suggested Valuation: Uses valid close prices of top candidates to estimate a valuation range and midpoint.

---

🖥️ Local Setup

1. Clone the Repository

git clone https://github.com/your-username/CompWise_AI.git
cd CompWise_AI

2. Install dependencies**

pip install -r requirements.txt

3.Set your OpenAI API key**

Make sure your OpenAI API key is exported as an environment variable:

export OPENAI_API_KEY=your-key-here

4.Run the app**

streamlit run app.py

📁 Files Overview

File Name	Description
app.py	Main Streamlit application
top3_gpt_explanations.csv	Ranked candidates with GPT-generated explanations
feedback_log.csv	Stores user responses on comparable property suggestions
requirements.txt	Python dependencies

---

📊 Tech Stack

• Python with Streamlit
• Altair for data visualization
• OpenAI GPT-3.5 for explanations
• Pandas for data manipulation

---

🙋‍♀️ Author

Built with ❤️ by Harman