German startup Zoomsquare has taken the first step towards artificial intelligence in property hunting, launching its fully functioning online chat advisor, zoomBot.
The chatbot is as an interface that sits between messaging and artificial intelligence and uses machine learning, natural language processing and semantic text analysis to help users find their dream home online.
The zoomBot must go through several stages before the final, fully-functional version can be unveiled. The newly-released beta version is a chat function platform, based on Facebook’s Messenger for desktop and mobile. The bot’s functions allow it to process questions asked and information given by users, and the end product will be able to search for suitable results and present them to the consumer.
“With this early stage beta version we are just at the beginning of everything,” said zoomsquare founder Andreas Langegger. “Our prime goal was to bring zoomBot to market as early as possible. We clearly see a very high potential in the bot technology but what’s even more important is that with every request the zoomBot improves his ability to recognise similarities, associations and to work with probabilities."
“Thanks to certain algorithms and machine learning the bot continuously improves itself.”
To access the zoomBot, users open the Facebook messenger application on a mobile phone and enter “zoomBot” in the search field. Desktop users have two choices: They can either use the messenger function at zoomsquare.de or visit www.facebook.com/zoomsquarebot.
The zoomBot is currently available in English and German, operating in Austria and Germany. A whatsapp version of the zoomBot is also in the works.
Domain.com.au has been experimenting with a chatbot via Facebook but Langegger says his firm’s product incorporates much more sophisticated natural language processing and approaches to artificial intelligence functions, and was very pleased to have launched a bot ahead of major competitors immobilienscout24, Rightmove, Zoopla and Zillow.
“Since the initial launch of the platform our entire system is based on semantic text analysis and machine learning,” he said.
“This means we analyse all indexed insertions including the long text for hundreds of characteristics. If the user now expresses certain or special requests we are able to respond very accurately to it. At the same time previous and similar inquiries and resulting probabilities are supporting a perfectly fitted matching.”
He said the introduction of the bot was a key move toward the company’s goals of providing a smarter way for consumers to search for property online.
“When we launched zoomsquare in October 2013 we had the vision about how great it would be to talk to zoomsquare just like you would be talking to a friend on where you want to live,” he said.
“That you could just tell the system your exact preferences in terms of apartment hunting. So for example if you could just say that you are looking for a refurbished apartment in a certain district with a bathing tub and a park for the dog close by, five minutes away from the next subway station.
“Right now we’ve already gotten a huge step closer to that goal.”
Zoomsquare’s tech savvy has lured a prominent investor in Austrian Hemann Hauser, a pioneer on the British startup scene and who co-founded Silicon Fen in Cambridge, known as the English Silicon Valley. Hauser sees the future in intelligent technologies.
“We won’t have zeros and ones anymore but probabilities which will take humans by the hand,” he said.