Feature: natural-language filters
“Under 100” is a filter, not a keyword.
Shoppers type what they want in one line. Gazelle reads price ranges, stock and sale requests out of the search, applies them as filters, and searches the rest.
Example searches
“headphones under 150”
Headphones priced up to 150
“between 20 and 50 mugs”
Mugs priced 20 to 50
“running shoes on sale”
Discounted running shoes
“1000 tl altı ayakkabı”
Shoes under 1000 (Turkish)
What it understands today
The filters show on the results page, so the shopper can see and remove them.
- Under / below / max a price
- Over / above / from a price
- Between two prices
- In stock
- On sale / discounted
What it doesn't do
This is rule-based parsing, not a chat assistant. It won't answer questions like “what's a good gift for a runner”. We'd rather say that than oversell it.
Questions
- Which languages are supported?
- English and Turkish phrasing for prices, stock and sale today.
- Does it use AI?
- No large language model is involved; parsing is fast, predictable and runs on every search.
Related
Free to start