Real Estate Research Tool
AI-powered research assistant for extracting insights from real estate news articles
Domain / Function
Generative AI • NLP • Information Retrieval
Project Overview
Developed an AI-powered research assistant that automates real estate market analysis by extracting insights from online news articles using a Retrieval-Augmented Generation (RAG) pipeline.
Users provide article URLs, and the system scrapes, cleans, embeds, and stores the content in a vector database. Natural language queries are answered using LLaMA 3 via Groq, ensuring accurate, source-linked responses.
Streamlit Research Interface
RAG Workflow
Streamlit Interface
Key Features
- Automated article scraping from user-provided URLs
- Semantic chunking and embedding generation
- Vector-based retrieval using ChromaDB
- Context-aware Q&A using LLaMA 3 (Groq API)
- Source-linked answers for transparency
- Interactive Streamlit web interface
Project Details
This project showcases strong expertise in Generative AI, NLP, and RAG architectures. By combining web scraping, vector databases, and large language models, the system enables fast, reliable, and scalable real estate research without manual reading.