CARY – Jordan Meyer, a data scientist who lives in Cary, is one of three members who are $1 million richer after winning a contest intended to improve the accuracy of a home value estimating program offered by real estate firm Zillow.

Jordan Meyer

Meyer confirmed to WRAL TechWire that he is a member of the winning team. According to his LinkedIn profile, is a data scientist at DataRobot. He’s a graduate of Appalachian State who also participated in UNC-Chapel Hill’s graduate program focusing on business analytics.

“I felt a combination of luck and pride,”  Luck says about his reaction to winning, “because a lot of circumstances
lined up for me to have the opportunity to compete. Pride, because our team worked very hard.”

Meyer teamed with Chahhou Mohamed of Morocco and Nima Shahbazi of Canada to form the winning team they nicknamed ChaNJestimate, which is a play off their first names. According to Zillow, the three met and decided to form a team while participating in the competition that last some two years and involved some 3,800 teams across 91 countries.

He says the prize money will be split three ways, and unlike some lottery winners he plans to stay at his current job.

“I plan to invest my share of the winnings and I’ll definitely keep working at DataRobot,” he explains.

Where to invest?

“In real estate.”

A video posted by Zillow shows a surprised Meyer being presented with a huge check by a Zillow representative.

The three rallied from third place at an earlier stage of the competition to win the top prize.

The Zillow program is called Zestimate, which helps home owners estimate the value of their homes. It already automates valuations of 110 million homes.

“Everyone knows the Zestimate, whether you’re a data scientist or not. It’s still unbelievable that we won the million-dollar prize,” said Meyer in a statement. “I feel so lucky and so proud of all our hard work.”

According to Zillow, which recently began offering all-cash offers to home owners in the Triangle, the three have yet to meet in person although they have collaborated online for “hundreds of hours” in developing the winning algorithm.

“Jordan’s specialty in developing deep neural network systems – which grew from an initial interest in biologically-inspired algorithms, or computer systems that mimic the neural circuitry of the brain – was a powerful addition to Chahhou’s neural network expertise and Nima’s strong track record of building accurate predictive models,” according to Zillow.

The three utilized” deep neural networks to directly estimate home values and remove outlier data points that fed into their algorithm,” according to Zillow.

“They also leveraged publicly-available, external data including rental rates, commute times, and home prices, among other types of contextual information, such as road noise – all variables that factor into a home’s estimated value.”

Utilizing machine learning, the three created a solution that reduced Zillow’s error rate by more than half a percentage point to less than 4 percent and beat a Zillow  benchmark by 13 percent.

Zillow Offers (all cash) launches in competitive Triangle real estate market

Zillow acknowledges that Zestimate is “$10,000 off of the actual sale price for a typical home” but with improvements made from information contributed during the competition the Seattle-based company believes future estimates could be more accurate up to $1,300.

“People are incredibly passionate about their home and understanding its value, and we are amazed by the winning team’s hard work the past two years to make the Zestimate even more precise,” said Stan Humphries, chief analytics officer and creator of the Zestimate, in the announcement.

“We’ve been on a 13-year journey making the Zestimate more accurate, and hosting Zillow Prize allowed us to invite thousands of brilliant data scientists from around the world to join us on this journey. We’re so proud that the winning team’s huge achievement, and the work of all the teams in the competition, will provide millions of homeowners with a better understanding of one of their biggest life investments.”

Participants in the contest used the online platform Kaggle.

Zestimate launched in 2006.