Md Ariful Islam
All projects

AI-Powered Real Estate & Legal Data Automation Pipeline

A Python pipeline that finds high-value real estate leads by cross-referencing public legal case records with MLS ownership data — using Playwright browser automation and LLM-assisted parsing to verify long-term property owners.

  • Python
  • Playwright
  • SeleniumBase
  • Groq API (LLM)
  • Pandas
  • lxml
  • Requests
View code
AI-Powered Real Estate & Legal Data Automation Pipeline
AI-Powered Real Estate & Legal Data Automation Pipeline screenshot

Overview

A Python-based automation pipeline that identifies high-value real estate leads by combining public legal records, search engine data, and MLS property databases. It automates the end-to-end workflow of discovering property owners involved in legal cases and verifying their ownership duration through MLS records — integrating browser automation, AI-assisted data parsing, and multi-source data collection to generate structured datasets for lead generation and market analysis.

Key Features

  • Playwright-powered browser automation for authenticated data extraction from the BrightMLS platform
  • AI-assisted data parsing using LLMs to extract structured address data from unstructured search engine results
  • Automated scraping of legal case records from the Bucks County ProPublic portal
  • Multi-source data integration combining legal records, search engines, and MLS property databases
  • Fault-tolerant automation pipeline with exception handling and long-running monitoring

How It Works

  1. Legal case scraping — Extracts legal case records and involved parties from the Bucks County ProPublic portal.
  2. Address discovery — Uses automated search engine queries and AI-assisted parsing to discover missing addresses for individuals listed in legal filings.
  3. Ownership verification — Logs into the BrightMLS platform using Playwright automation to verify property ownership duration and ownership history.
  4. Data structuring — Normalizes addresses and compiles verified datasets containing owner information, property details, and ownership timelines.

Tech Stack

Python · Playwright · SeleniumBase (undetected Chrome) · Requests · lxml · Pandas · Groq API (LLM) · DuckDuckGo Search · Google Search · usaddress

Project Structure

project/
├── main_scraper.py       # Pipeline orchestrator
├── bright_mls.py         # Playwright automation for BrightMLS ownership verification
├── search_address.py     # AI-assisted address discovery and parsing
├── filters.config        # Configuration for credentials and filters
├── zipcodes.csv          # Geographic filters for targeted ZIP codes
└── readme.md

Output

Generates a structured dataset (output.csv) containing owner names, verified property addresses, ownership duration, legal case references, and representative contact information — used for real estate lead generation, market analysis, and legal data insights.