Natural language search engine
Natural language real estate listing search
Project
For the startup GoFlint, a natural language search engine was designed to transform the user experience. This innovative solution allows users to express their real estate needs in their own words, without relying on complex filters. The AI analyzes the query and searches the national database of listings to identify the most relevant results. The solution is now accessible via WhatsApp.
The need
On a listings portal, search goes through filters: city, surface area, budget, number of rooms. A real request is more nuanced, and many criteria do not fit in any box.
How it works
Users describe the property they are looking for in their own words. Semantic search matches this request with the listings in the national database, even when the wording differs. On WhatsApp, the assistant runs the search within the conversation and creates an alert for every new listing.
What the project shows
Natural language search applies to any rich catalogue: listings, products, job offers, document collections. It replaces the filter step with a simple sentence.