Frequently asked questions
Short answers to what people ask most. Each question has its own page with the full explanation.
How does NeuralSearch find listings?
It runs keyword, semantic and image retrieval at the same time, merges the three result sets, then re-scores the merged shortlist against the specific constraints in your question. Retrieval decides what is considered; the re-scoring step decides what you see first.
Can I search in plain language?
Yes — full sentences are the intended input, not a fallback. "A low-mileage estate under 20,000 with a tow bar" works as written; the constraints are read out of the sentence rather than typed into filter boxes.
What is image-aware ranking?
The photographs on a listing are read as a signal in their own right, not just its text. That lets a query like "bright, with a lot of natural light" match listings whose descriptions never use those words, and it decides which photo is shown first.
Which sources are indexed?
Listings are collected from public listing sites and normalised into a single shape, so one search covers catalogues that describe the same thing in different words. Every listing keeps a link back to its original source.
How fresh are the listings?
Listings are re-collected on a schedule and each one carries the time it was last seen. Availability and price should always be confirmed on the original source, which we link from every listing.
Is NeuralSearch available in Polish?
Yes. The interface is available in English and Polish, and you can search in either — the query is understood in the language you write it, not translated into English first.
Is my search history private?
Searches and conversations are tied to your session so you can return to them, and are not shared with other users. Analytics are off until you accept analytics cookies, and are aggregate rather than a record of individual people.