How AI is helping NoBroker find you a house closer to work

Akhil Gupta, the founder of NoBroker.com, says it is probably the only property listing portal that has does not employ field agents. The reason is simple: technology.

“We rely on data which we collect from different sources and use technology to draw insights from them to come out with the best possible outcomes for our prospective customers,” Gupta, who is also the real estate platform’s chief technology officer, told TechCircle.

Bengaluru-based NoBroker, which does not charge brokerage, has been deploying newer technologies such as artificial intelligence and machine learning to not only better its product offerings but also to expand while keeping employee costs in check.

Gupta points to the portal’s most basic backend product, which it internally calls the price predictor.

As the name suggests, the predictor is responsible for answering queries about rent depending on input criteria such as place or size of the flat in a particular region.

“Before it delivers a conclusion, the predictor is fed with two scores – commute and living. These scores themselves run on individual AI and ML algorithms after considering the data it has been trained on,” said Gupta, who was previously an application engineer at Oracle and technical lead at PeopleFluent, a software-as-a-service talent management solutions company.

The commute score represents the connectivity of the city to the region where the property is located and the living score considers aspects such as size and amenities around the property.

Gupta said that these algorithms were being trained on both data that was being crowdsourced and information collected over the years by listing properties and providing other related services such as rent payments and agreements. 

“Most of our crowdsourced data that helps us get to commute and living scores come from APIs of Google Maps, Ola and Uber,” Gupta said.

According to Gupta, the predictor helps NoBroker showcase property prices that matches consumers’ needs and is based on a scientific method. 

Incidentally, home rental startup NestAway also uses an AI engine built in-house to predict rent.

Gupta also said that the company has another product called ‘travel-time search’. 

It runs a series of algorithms to check periphery and traffic to let the customer know how long it would take to travel between his/her office and the property.

“Most of our customers, and customers in general, don’t want to stay far away from work and this is why we came up with this product,” Gupta said.

The product has a few more bells and whistles.

“What the consumer can do is let us know where the office is and how much travel time he expects. Based on this information we will suggest a property in a particular radius which might not be the closest, but surely will have the shortest commute time,” he explained. 

Founded in 2013 by Indian Institute of Technology (IIT) graduates Gupta, Amit Agarwal and Saurabh Garg, NoBroker currently has more than 400 employees. Around 50 of these handle technology operations.

Gupta said technology has been having an instant impact in terms of more customers visiting the portal and the firm has also been able to expand without having to hire too many employees.

“When our rivals were at the stage of 10,000 transactions per month, they had a total employee strength crossing the 2,500 mark. We have used technology to cut back on our HR expenses,” Gupta said.

NoBroker, which works on a subscription and referral model, is planning to expand to other cities within the next six months.

The company’s strategy of expansion into different regions or cities starts with offering rent payment and rental agreement services.

“This helps us in understanding the lay of the land, the transaction sizes etc,” he explained, adding that the portal was currently recording 10,000 transactions every month.

In an earlier interaction, Gupta had said that NoBroker was adding one lakh customers every month.