Create parsed job records from hourly site scrape

You run hourly site scrapes that return raw HTML, leaving new job postings unstructured and hard to triage. This extracts the first listing into a clean job record so coordinators can review it the same day.