With natural language search, is the AI component simply using the sentence input to create filters? Or has it become more complex since this video was released? If my search term was iPhone 16 Pro 256GB, would it know to only return that exact model, or will it also return iPhone 16 Pro MAX 256GB as well?

This demo was created before NL search was a feature. But the LLM will use the sentence to generate the filters. You can give it more concrete instructions with the system prompt as well

Do you have any examples?

Of NL search or using a system prompt?

Right now I'm doing a loose search on typesense, then passing all of the results to OpenAI and asking it to match the products for me

Could you check this out: https://typesense.org/docs/29.0/api/natural-language-search.html#create-a-natural-language-search-model This is a guide for NL search. I'll update the demo to use the Typesense native NL search as well

ok thanks