Tailored Ads and Controlled Photos: How gpt4 is changing real estate
[ad_1]
Artificial intelligence is revolutionizing even one of the most reluctant to change sectors: real estate. With this technology you can predict the value of a propertyyou can analyze the images of the houses to obtain information on their contents, make them anonymous and detect material to share the images securely with potential buyers.
They know it well proptech startupwhich use machine and deep learning mechanisms in automatic property valuation models, to create the perfect match between home sellers and home buyers, for the automatic analysis of videos, photos and floor plans, in models to make estimates and predict sales times in different areas of each city.
These are machine learning models that learn on the basis of the feedback they receive: on the characteristics of a property, for example, the model estimates the value of the house. But what do they learn from?
“Milan, for example, carries out 20-25 thousand residential real estate transactions a year – he explains Fausto Maglia, CPO of Casavo, next generation proptech platform for home buying and selling -. We train our algorithm on a quarterly basis: we see what it has estimated and we give it the real sales figure so as to correct itself in case of rejects. It’s about a supervised training process of the model”.
Startups
Casavo fires 30% of employees. Tinacci: “Difficult but necessary choice”
by Italian Tech editorial staff
The Netflix of homes
As Netflix suggests movies and series based on what we’ve already watched, Casavo does with real estate ads. “We recently developed Seller Buyer Match – says Maglia – unlike the ad platforms which are digital catalogues, us we work to generate the registration of a contact between seller and customerfor this we are working on a recommendation system a la Netflix. We also give visibility to new announcements from the seller who is still evaluating whether to sell a house”.
Another use of artificial intelligence is that of image recognition:
“We are developing deep learning and computer vision capabilities to understand what user photos are referring to.” It’s about computer vision tools that allow you to check the uploaded images: if photos of the windows are missing, the intelligent algorithm signals it and is also able to recognize if there are people in the images and obscure their faces or anonymize the floor plans.
Real estate
“A great digital enterprise can also be done in Italy”: Casavo closes a 100 million round
by Eleonora Chioda
Personalized ads
Until today, the announcements of houses for sale were the same for everyone. Now, with artificial intelligence, the aim is to create ads that are “tailored” to the customer. “His needs will guide the announcement – concludes the expert -. Schools and parks in the area will be indicated to a family with children, while gyms and sports centers will be indicated to a sportsman and so on. We will use gpt 4 but we will give the parameters with our usual spasmodic attention to privacy”.
[ad_2]
Source link
