I wonder how this compares to algolia? I recently added that to an open source project I released recently and the search results aren’t just not relevant they aren’t even close to correct. It works by first indexing your site by scraping it.

Not sure why everyone seems to use it.