Currently, a new connection is established every time we scrape Youtube. Implement connection pooling (inbuilt in `requests`).