Real Estate Lead Bot: Indiana Court Records to Clean Leads
Anas Yaseen
One search on Indiana's court website took a real estate team 20+ minutes and three different tools, many times a day.
The problem
The team finds property leads in public court cases. Octoparse copied 1,000+ case pages but broke on missing fields and put names, streets, cities and zip codes into one messy column. So they split it with DeepSeek and filtered it in Google Sheets before it reached the CRM.
What I built
A simple dashboard the team runs themselves:
Enter the search and click Run
Keep only the case types they want
Watch the progress live
Download a clean CSV, ready for the CRM
The hard part
Court records are messy: a lawyer's name inside the address, five defendants on one case. Every clean-up rule the team used to apply by hand is now built in. A browser was too slow, and the site blocks busy visitors and only works from the US. So the bot reads the site's own data directly, like a normal visitor, through rotating US connections.
Results
20+ minutes per search down to about 1 minute
50 court cases turned into 77 leads in 31 seconds
No more Octoparse, DeepSeek or Sheets clean-up
5-star review: "He is amazing! I would work with him again!"
Tech: Python, curl_cffi, BrowserForge, Streamlit, proxy rotation
Full case study: https://mranas.dev/work/court-records-lead-bot
Like this project
Posted Oct 19, 2026
A real estate team spent 20+ minutes per search pulling leads from Indiana's court website with three different tools. Now they enter the search and click Run. The bot cleans names and addresses and saves a CRM-ready CSV in about a minute.
