I have a roster of 705 companies that appear in both the Australian and UK Modern Slavery registries. The task is to build a repeatable Python workflow that: • Automatically locates and downloads each firm’s latest Modern Slavery (MS) statement from both registries, saving the original files locally for audit-trail purposes. • Parses the text, then scores the content against the Ethical Trading Initiative (ETI) framework so every criterion becomes a clean numeric or categorical field. • Adds a second-stage variable that flags and classifies any MS-related allegations—sources for those allegations will be provided from my existing database. This variable is the priority focus and should be robust enough to capture multiple allegations per company when they exist. Deliverables • Well-commented Python scripts (requests/BeautifulSoup or Selenium for scraping; pandas/numpy for cleaning and scoring; any NLP library you find efficient). • An Excel workbook containing one row per company and only data columns—no charts or visuals. • A concise methodology document describing the step-by-step Python process you followed and the data-extraction techniques employed, so the study is fully reproducible. Acceptance A sample of 50 randomly selected companies must run end-to-end without manual intervention, populate every ETI data point, and correctly reflect the allegations variable before we sign off on the remainder of the list.