Web Scraping / Data Extraction Developer – Company Facility Discovery

We are looking for a developer to build an automated system that identifies manufacturing plants, warehouses, distribution centers, and fulfillment centers operated by a given company. The system will receive a company name and domain, then: * Search the company’s official website, subdomains, supplier portals, and career pages using a consistent set of facility-related searches. * Discover relevant location directories, supplier documents, routing guides, facility maps, PDFs, spreadsheets, and job postings. * Extract facility names, addresses, cities, and facility types. * Preserve the exact source URL and supporting evidence for every location. * Exclude offices, retail stores, R&D sites, and other non-shipping locations. * Stop after a defined number of official sources rather than conducting open-ended research. The solution should use reusable extraction methods for: * Standard HTML pages * JavaScript-rendered pages * Interactive maps and underlying APIs * PDFs and spreadsheets * Images or hover-card graphics * Career portals and job postings This must be a company-agnostic system—not custom-coded separately for every company. Individual companies will be used as test cases to improve the general process. Requirements * Strong experience with Python and web scraping * Experience with Playwright, Selenium, Scrapy, or similar tools * Ability to inspect APIs, JavaScript applications, and embedded map data * Experience extracting structured data from PDFs and spreadsheets * Experience using LLMs for classification and evidence extraction * Strong data-normalization and duplicate-matching skills * Ability to design clear confidence, review, and failure statuses * Must provide readable, documented, maintainable code Accuracy is more important than finding every possible location. The goal is to produce a substantial list of highly reliable, officially supported facilities—not claim perfect coverage or use questionable third-party sources. Please describe similar extraction systems you have built and how you would approach this project without creating company-specific scrapers.