darren whittingham - Fotolia

At AT&T, CDO responsibilities to include all things AI

At most companies, the chief data officer role tends to focus on data governance and management issues, but at AT&T, AI is set to be a big part of the executive's remit.

Many executives stepping into a newly created chief data officer role might start their time in the position with a data governance project or analytics software rollout.

But for Steve Stine, who recently took on CDO responsibilities for telecommunications company AT&T, it's all about AI-powered automation.

"We believe that there are many capabilities that are ready for prime time and some that will continue to evolve," Stine said. "Bottom line is, yes, we're a believer in AI."

CDO responsibilities have traditionally focused primarily on data management issues. Things like how data gets stored and used as a corporate asset have dominated the time of typical CDOs. But AI is opening up new business opportunities and, at least for now, most enterprises lack a single executive or business unit responsible for overseeing how AI is implemented and used.

Stine is using this responsibility gap to make AI and automation a big part of his remit as AT&T's CDO. The decision makes sense given his background. Prior to becoming AT&T's first CDO, Stine worked on automation projects related to customer service in his role as senior vice president of technology planning and optimization.

Rolling out AI enterprise-wide

Now he's looking to apply AI and automation throughout the company. For example, he and his team are developing and will soon deploy chatbots to support field technicians. The techs will interact with the chatbots through a company-issued tablet, enabling them to retrieve billing information, make account changes, activate television set-top boxes and get answers to technical questions.

Previously, field techs had to do all this by calling into a live person manning a support line.

Steve Stine, AT&TSteve Stine

The platform is being built using natural language processing and machine learning models to improve performance over time. The project is currently in proof-of-concept testing, and Stine's team expects to have it go live in 2018.

Stine's team is also developing machine learning models to detect service interruptions using network data and to dispatch field technicians and give them optimal routes based on service requests, traffic and weather data.

"When you start adding those things together, it provides insights that we weren't previously aware of," Stine said.

The primary theme of all these projects is automation using machine learning and natural language processing to do jobs that were previously performed by people. Some see this trend toward automation in the workforce as a threat to jobs, but Stine sees it differently. He said that instead of taking people's jobs away, automation at AT&T is enabling workers to skip the drudge work and focus on more interesting tasks.

For example, the company has already done a lot around process automation. Teams including human resources, sales, finance and IT have all had bots developed to perform some task related to their work. This typically involves simple processes, like moving data from one application to another or pulling data from web forms into a spreadsheet.

These simple and repetitive tasks were once performed by people, but are now done by bots. But the people haven't been eliminated from the process. Now, instead of performing the tasks directly, the workers are overseeing the bots in some cases and, in others, they are working with the data gathered by the bots to develop deeper insights.

"It changed people's jobs, but it didn't eliminate them," Stine said.

Stine sees hope behind AI and automation

This is why Stine is optimistic about the effects of automation on the future of work. There's been no shortage of hand-wringing about the potential threat to jobs posed by AI. Some people today are even talking about how AI will automate half of all jobs. But Stine sees it differently.

"We want to repurpose people for more meaningful tasks," he said. "I'm optimistic about the effects of automation on the workforce."

Primarily, he sees people moving from repetitive tasks to more insight-driven work. Because of this, he sees part of his CDO responsibilities as getting people self-service tools so they can take advantage of the new opportunities to use the data to become more productive.

Stine declined to comment on specific software and technology platforms currently used at AT&T, but said the company is constantly evaluating new tools, particularly when it comes to AI tools and techniques.

"The components that make up AI can make it valuable," he said. "We keep a really close eye on it. There is a lot of investment and interest, so we believe there will be a quickening of the pace of development."

Next Steps

Find the right fit for your CDO position

Chief data officer jobs continue to grow in enterprises

City of Boston's incoming data chief looks to grow the role

Avoiding inflated expectations is key when implementing AI

Dig Deeper on Careers in artificial intelligence