AI and machine learning promise to not only improve the customer experience, but also change the way companies operate. For this reason, enterprises should consider integrating these technologies into digital transformation plans to stay competitive.
By 2019, 40 percent of all digital transformation initiatives will be supported by cognitive/AI capabilities, according to IDC.
"The time for AI has now finally come, because of the technique of deep reasoning connected with the amazing amounts of computer power and data," said Sanjay Srivastava, chief digital officer of Genpact. "We're seeing the impact of that in our personal lives already. The opportunity now is to apply those technologies in the business context."
For example, we can now use AI for account management and customer service systems across industries. "The benefit isn't just the fact that you get productivity, but the fact that you can scale very quickly," Srivastava said.
Gartner predicted that by 2018, 20 percent of business content (such as shareholder reports, legal documents, and press releases) will be authored by machines. And IDC expects that at least 20 percent of all workers will use automated assistance technologies by that year.
Tools like IBM Watson "introduce machine learning to understand what human beings have a hard time seeing in the context of their work," said Brian Solis, principal analyst at Altimeter. "Machine learning will allow companies to see things they wouldn't otherwise, because of the cognitive bias that exists in the relationship between humans and the data they collect."
These tools can help companies learn in ways that accelerate innovation, Solis said. "AI and machine learning will be really helpful tools in companies getting closer to customers and objectives," he said. If you better understand the context of the customer's world, you can better help them accomplish their goals, and base digital transformation efforts on that information. "Use those insights to reverse-engineer new ways to apply AI to create value," Solis said.
Companies can also use AI internally to study productivity and employee engagement. "Employee experience is the next customer experience," Solis said. "A lot of that is gaining insight into ways to introduce operational models they couldn't see before."
The automation factor
While some studies predict that nearly half of today's jobs could be replaced by a robot within 20 years, automation can ultimately aid productivity and digital efforts, according to Marc Cecere, vice president and principal analyst on Forrester's CIO role team.
"Nearly all processes within an IT org will experience some automation," Cecere said. Some of these will be replaced entirely, he said. Tasks such as backups, job scheduling, password resets are all in the process of being automated.
"More insulated from automation are processes like agile development, which requires a high level of creativity, but even higher levels of social interaction and intelligence, as business and IT people need to interact closely to create something new," Cecere said.
AI and machine learning can also help determine which business practices can be automated, Cecere said. "Machine learning, in particular, can be used to rapidly create scenarios of many practices and match them with the data to see which combinations of practices are optimal," he added.
Increased automation is beneficial when it helps the most creative people in a company be more productive, said Andrew Moore, dean of the school of computer science at Carnegie Mellon University.
For example, AI and machine learning could be useful in his job as a dean of a college because he could ask a platform, "Are faculty course evaluations getting better or worse in the past few years?" and have a machine pull that data up automatically.
"It would be nice to get quick insights from your data," Moore said. "It's the same as installing digitization software. Except this is going a bit further, to the point where you can ask questions and the AI system behind the scenes can figure out what data sources are relevant to use, and the results."
A number of companies are already using AI and machine learning tools in digital transformation efforts. Health insurance provider Humana recently deployed an AI tool to aid call center employees and customers.
"The Cogito AI tool helps our agents recognize when a conversation is hitting a snag, slowing down, or escalating into frustration," said Geeta Wilson, director of Humana's Service Experience of the Future. "The agents are so busy trying to solve the customer's problem that they can miss subtle cues in the conversation. That's when the tool alerts them, in real-time, about how they can improve the call and course-correct while they are still interacting with the customer. At that point, they can recover the call and turn it positive again."
Humana has already seen improved issue resolution scores and net promoter scores, Wilson said.
Fashion retailer Obsessory.com uses a deep-learning algorithm to review and fine-tune customer choices on its website. "Advanced artificial intelligence helps our business make sense of a huge data set by going through millions of records," said co-founder Farnoush Mirmoeini. "For us, searching through millions of products and images is not humanly possible, and having algorithms to mimic what humans do and how they think and perceive things is key to our value proposition."
And the Treasure Island Las Vegas casino uses IBM Watson in its mobile guest app, so customers can request a towel or a restaurant recommendation and receive answers via SMS messages.
"Everything from machine learning to AI to the chatbot phenomenon to robotics is starting to come to a place in the market where the use cases are real, the benefit value realization is real, and you can do it at a price point that is very interesting to a lot of companies," said David Nichols, a partner at Ernst & Young.
"You don't have to wait -- you need to do this right now," Nichols added. "Some companies are sprinting into this and getting great results, while others will pay a competitive price."
Where to start with integrating AI and machine learning into digital transformation efforts? First, research different approaches to implementing AI and machine learning solutions, and determine which would work for your company, the experts said.
You also need to ensure that you have access to large datasets to power the AI, Srivastava said. Then, it's about building the algorithm.
Unlike in previous years, when automation was largely owned by proprietary technologies, today most of the AI innovation is happening in the open source world, said Genpact's Srivastava. "Don't sit around waiting for a large software company to deliver the ultimate solution," he said. "It won't happen."
Of the mainstream digital assistants, Amazon's Alexa currently has the best open software development platform, according to Moore. If you are a consumer-facing company, Moore recommends assigning an IT employee to implement Alexa Skills, to begin integrating with the system.
These technologies can change your business operations by bringing value to customers, so they start using your services differently, Srivastava said.
For example, you can tell Alexa, "Get me an Uber to the airport on business expense." That could involve Alexa logging into Uber, ordering the car, sending an email to a work account, which forwards it to an expense management system, which approves it and reimburses the employee -- all without picking up a phone, taking out a credit card, or filing a report.
"Just in that example, four or five different AI techniques are applied, and so changed the business," Srivastava said. "There is an amazing opportunity to apply these changes to business concepts on the back of AI technologies."
Most CIOs are caught up in driving operational efficiency, and are burdened with legacy systems, Srivastava said. Still, "they have to take time out and get onto the AI curve, because it's significant in the impact it's going to have on your business," he said.