Deep learning expands on machine learning by allowing intermediate representations of data to solve complex, data-rich business problems.
No Jitter, a website providing daily commentary and analysis of enterprise IP telephony, unified communications, and converged networking, proclaimed recently: “If you had to choose a subject that emerged as a unifying theme across the contact centre market in 2017, you’d have to go with artificial intelligence (AI). … In 2017 we saw a groundswell: multiple contact centre software vendors announcing, incorporating, trialling, and sometimes delivering chatbot, predictive routing, and other types of machine learning and natural language understanding.”
AI in the contact centre
Chatbots might represent the first, and currently the largest, manifestation of AI in the contact centre, but they are certainly not the only one, and a focus on chatbots tends to draw attention from other potentially transformative applications – machine learning in particular.
For example, machine learning has been employed to develop algorithms that can predict trends in agent attrition and absenteeism, based on established behavioural patterns, employee actions, and data compiled from a multitude of sources within your contact center infrastructure.
Algorithms can also help contact centre managers understand when agents are performing at their best, enabling them to optimise shifts and rostering.
Beyond AI: machine learning
Paul Stockford, chief analyst at Saddletree Research — a company providing market research and intelligence to the contact centre and customer care industries — is somewhat sceptical about AI delivering automated responses to customer enquiries, a key role for chatbots.
“It’s easy to see how the industry has been captivated by the allure of AI,” he says. “But it’s important to maintain a realistic perspective regarding AI in the contact centre. It is in its earliest iteration and while AI has enormous potential, don’t get ahead of yourself.”
However, he explains the distinction between AI and machine learning and is a big fan of the latter. He says it “gives a computer the ability to learn without being programmed to learn … in order to make data-driven predictions or decisions.”
This post from Network World, a website specialising in news, intelligence, and insight on networking and data centres, agrees with the benefits of machine learning in transforming customer service. Its writer contends that machine learning “has the potential to revolutionise the way your company does business, but only when it’s used correctly.”
“When used properly, machine learning will take high volumes of customer interactions, which used to slow down companies with greater workloads, and use them to find useful trends,” he says. “As algorithms and human-replacing software proliferate more and more, they will only become cheaper and easier to use.”
He says companies hoping to stay at the top of their game should “recognise the forthcoming importance of machine learning, and take steps to integrate it into their business models for future success.”
But what about deep learning?
Though it’s still early days for the exploitation of machine learning in support of customer service, another technology is already on the horizon: deep learning.
According to Alexander Linden, research vice president at Gartner, deep learning expands on machine learning by allowing intermediate representations of the data to solve complex, data-rich business problems.
“Deep learning can, for example, give promising results when interpreting medical images in order to diagnose cancer early. It can also help improve the sight of visually impaired people, control self-driving vehicles, or recognise and understand a specific person’s speech,” he said.
Gartner predicts 80 percent of data scientists will have deep learning in their toolkits this year. They say deep learning delivers superior data fusion capabilities over other machine learning approaches and predict that, by 2019, deep learning will be a critical driver for best-in-class performance for demand, fraud and failure predictions.
According to Mikhail Naumov, co-founder and president of AI technology company, DigitalGenius, the contact centres best able to make use of deep learning will be those with large volumes of historical customer service data, such as email transcripts and chat logs.