AI Website Classification Pipeline (Python + ChatGPT)

Delivered a clean Python script for a one-time local run on Windows to process a large list of domains at speed. The script visits each homepage, extracts core content signals including page title, meta description, first H1, and a short text snippet, then sends the structured payload to the ChatGPT API using client-provided classification instructions. Final classifications and scraped fields are exported to CSV for immediate analysis. The implementation uses requests, BeautifulSoup, and lightweight concurrency to improve throughput while keeping setup simple.