New Mobile Solution Uses AI to Deliver a Personalized, Hyperlocal Weather Experience

By Elissa Gootman

As a second-generation meteorologist, Kevin Eubank is familiar with a fear rippling through his profession: that technology is bound to replace human weather forecasters.   

Instead of running away from the latest technological advances, Eubank, who is Chief Meteorologist at the Salt Lake City NBC affiliate KSL-TV, believes the best approach is to embrace them—and find other ways to add value by providing uniquely human advice and insights.

Under Eubank’s leadership, KSL-TV has become the first television station nationwide to introduce Weather InSight—IBM and The Weather Company’s latest mobile weather solution. Now integrated in the KSL Weather mobile app, Weather InSight uses AI to give users hyperlocal, personalized weather information, in a way that’s highly intuitive and relevant to their lives.

“This is one of the coolest products The Weather Company has come out with, and it’s unlike anything else that’s out on the market,” Eubank said. “It learns where you are, what you’re used to, what you care about, how you like your information delivered. And then it serves up those items to you with intelligence. It filters out the things you aren’t interested in, and it filters up the things that you are.”

 

Kevin Eubank introducing viewers to the new AI-powered KSL Weather app, which highlights unusual weather developments without making users "data mine."

 

AI is built into the experience, allowing users to get the information in a way that helps them understand the weather. As a result of the new capabilities, KSL expects to see a 20 to 25 percent increase in ad revenue from its weather app.

“Our weather app users are savvy, and they want intelligent weather data,” Eubank said. “Our advertisers are equally as savvy, and they want a way to connect with that audience.”

A Personalized, Hyperlocal Approach

With thousands of weather apps on the market, it can be difficult to stand out. The team at The Weather Company that set out to build Weather InSight knew they needed to deliver personalized, hyperlocal information in an easily digestible format.

“The Weather Company has produced more than 600 weather apps for its U.S. media clients, but when we developed Weather InSight, we started from scratch,”said Rodney Thompson, Senior Digital Strategist at The Weather Company, an IBM Business.

Weather apps generally make users “data mine,” Thompson said. Users must continually click and scroll to get to the information they want, swiping past potentially irrelevant or confusing data like “dew point.”

“Data mining is so five years ago,” he said. “We’re getting used to a world where the right information just comes to us.”

Weather InSight is designed to free people from having to hunt down the information they care most about. Over time, it tailors the information presented to each user by cultivating an understanding of their concerns, interests and habits. It also highlights and draws the user’s eye to the most impactful weather information.

 

 

“The biggest complaint I get,’’ Eubank said, “is, ‘Hey, you never talk about my town. I want to know how much rain I have or what the temperature is going to be like,’ or, ‘It’s windy in my town, but you never talk about that.’ It’s a huge challenge for us, and the latest version of our weather app completely changes that.”

Eubank forecasts for Utah as well as portions of Arizona, Idaho, Nevada, Colorado and Wyoming. For the app, he can now create more specific, hyperlocal videos, with the knowledge that only viewers in the affected areas will see them.
The app exemplifies how AI can enhance the work of human specialists.  

“Kevin can’t get to every location in the Wasatch mountains and highlight certain things, and un-highlight certain things, and take things off the screen when they don’t affect people in a certain area,’’ Thompson said. “It’s literally humanly impossible to do all of that at a hyperlocal level. Where he adds value is coming in and breaking down the weather situation for you, and really making you understand the weather you’re experiencing.”

Built-In Privacy

Weather InSight was designed with a privacy-first approach. AI is built directly into the mobile app instead of operating at the level of the network server computer.

“The app is making observations and requesting the server for certain screens," Thompson explained. "But the server itself has no information about who you are or your device ID. Even though we’re really getting to know you and tailoring your experience, we know nothing about you—which in the world of privacy is super important.”

Weather InSight can be embedded in any app, including dedicated weather apps and other apps that might include weather information so users check them more frequently. It can also be branded differently for each client; when users open the KSL Weather app, Weather InSight appears as a round button at the top right of the screen, labeled “KSL In Depth.”

 

Using Weather InSight, the KSL Weather app presents the most relevant hyperlocal data.

 

Eubank, whose father owned and operated a successful weather company pre-internet, remembers being taught early that weather technology will constantly progress—and that part of the challenge of meteorology is continually identifying new ways to contribute.

“As a meteorologist, you have to find new ways to humanize the data,” he said. “AI refines that ‘what,’ but it will never be able to explain the ‘why.’”