Python Web Scraper / Data Engineer – County Recorder Tax Liens (Nationwide) United States

We are looking for an expert Data Engineer/Web Scraper to build and execute a scalable data collection pipeline to extract historical Federal Tax Lien records directly from county public records portals across the United States. Objective: Extract approximately 1,000,000 historical records from the last 10 years (2016–2026). The final dataset must filter for Active (non-released) liens with an initial assessment value greater than $5,000, covering all US states except Oregon (OR).Technical Reality Check:We understand that there is no centralized database and that many counties use aggressive anti-bot protection (Cloudflare, CAPTCHAs, paywalls). We do not expect you to scrape paywalled or offline counties. Your goal is to map out and scrape the top 150–200 highest-volume, open-access metropolitan counties in the US to help us aggregate our target volume. Key Responsibilities: Architect modular Python scrapers using Playwright, Puppeteer, or Scrapy to navigate county Recorder of Deeds/Clerk search portals.Implement robust anti-blocking measures including stealth plugins, human emulation delays, and rotating residential proxy networks.Build a parsing pipeline using BeautifulSoup/Pandas to normalize disparate county data structures into a unified schema.Cross-reference indexing data where possible to filter out liens that have a corresponding "Release" or "Withdrawal" instrument recorded. Required Schema Fields: Standard 5-digit County Name & State Taxpayer/Debtor Name Mailing or Property Address (where available) Recording Date (YYYY-MM-DD) Original Assessment/Lien Amount ($5,000+) Candidate Requirements: Proven track record scraping highly dynamic, JavaScript-heavy government portals.Experience handling Cloudflare/reCAPTCHA bypasses programmatically. Strong SQL or data normalization skills. To apply: Please state your experience dealing with county-level court or property databases and outline your preferred proxy/stealth configuration strategy for avoiding IP blocks.