There might be a lot of hype surrounding artificial intelligence (AI) and machine learning (ML). But communications provider AT&T is a believer -- and is transforming its business with the help of these technologies.
"Our mission at the chief data office is to data power AT&T through data insights and evolving technologies such as AI and ML," said Kim Keating, vice president of data science at the company. "This allows us to leverage vast amounts of data to provide insights and answers to critical business questions" affecting the company.
One are where the company is leveraging these technologies is in selecting retail store formats. By the end of 2019, it's planning to add 1,000 new retail points of presence. These will include pop-ups, mobile stores, traditional stores, and authorized retailers.
The data science team at AT&T is helping the company make those placement decisions using algorithms to help with site selection and store type, Keating said.
"Brick and mortar will still exist, but the standard will be different," Keating said. "With cutting edge analytics, we are learning that different markets are best served by different types of stores based on the needs of that market. As an example, we are taking into account data points such as growth communities and drive/walk time."
Kiosks are a newly launched format, primarily to dispense SIM cards to customers. The data science team is analyzing data about locations, mobile transactions, crime, and other areas to determine where these kiosks can best be placed.
The company is also leveraging AI and ML to improve efforts in forecasting and capacity planning with the dispatch field services that help customers every day.
"Our goal has been to optimize schedules for technicians, to get as many jobs done during the workday as possible by minimizing drive time between jobs while maximizing jobs completed per technician," Keating said. "The effects are positive for technicians, specialists, and customers."
With improved forecasting and capacity planning in its field dispatch efforts, AT&T has seen a 7% reduction in miles traveled per dispatch and a 5% increase in productivity.
Another use case is end-to-end incident management. AT&T uses an ML program to detect network issues as they first occur, not after they've been called in. That way customers don't even know there's been a problem.
"We sort through thousands of data points per second, finding patterns that indicate real-time issues," Keating said. "We can take seconds, not hours, to determine the smartest way to address the incident, impacting the fewest customers for the least amount of time."
The company is now able to manage 15 million alarms per day, a task that Keating said would otherwise be impossible.
"This technical expertise directly impacts our business, protecting several of our business services, consumer services, and infrastructure components," Keating said.
The opportunities to apply these newer technologies to solve business problems are limitless, Keating said. "We believe AI has incredible transformational power for any company, and we know it will digitally disrupt every industry—so you don't want to be left behind."