To what extent can businesses use contact centre artificial intelligence (AI) solutions as customer service agents in their call centres? The answer is that new, advanced systems can go a long way towards fulfilling many customer service requirements.

Chances are that your company has already taken some steps in this direction, both to free up human agents to deal with the most complex and challenging customer interactions and also take on new responsibilities. Forrester, for example, has noted that human agents may become super-agents to handle critical customer interactions that require deep subject matter expertise. AI also improves operational efficiency and reduces costs.

There is a new urgency these days that is driving companies to look to AI-powered chatbots, voicebots, virtual assistants and other solutions to carry much of the customer service load. The global health crisis is affecting human agents’ ability to provide customer service.

It has taken a toll on the contact centres in the Philippines. The country is the contact centre capital of the world with more than a million agents providing outsourced services to companies around the globe. Many of these contact centres went offline after the Philippine government imposed quarantine regulations, obliging companies to issue warnings of delays to customer service. Could this have been mitigated with some of the newer, advanced systems available?

Broadly, contact agents fulfil three roles: They provide information or initiate action in response to customer requests; they respond appropriately to customer emotions (for example mollifying an angry or frustrated customer); and they represent the “human face” of an organisation. Today’s smart contact centre technology can largely pick up these roles.

The most established customer service AI technology is the chatbot, technology that is able to interpret customers’ text queries, provide an appropriate response and initiate the correct actions within a company’s IT systems, such as changing a flight booking.

Amazon Web Services, for example, provides a complete suite of AI technologies that is able to accept verbal customer queries, convert these to text, translate them to another language, analyse them, initiate actions in AWS-hosted apps such as SAP and Salesforce, and then provide an appropriate verbal response to the customer.

Meet the human face of AI

Today AI is also being deployed to put a human face onto chatbots. Literally.

Digitally created human faces “speak” to customers rather than work through text interactions. They are able to show emotion through facial expressions and also respond using the correct tone for the conversation. Two New Zealand companies, Soul Machines and UneeQ, are among the providers of this technology. Their digital humans have been deployed in customer service roles by a string of large organisations, including ANZ Bank, Daimler, IBM, UBS and Vodafone.

Digital humans are in their infancy and you might find them a little creepy, but they are likely to improve rapidly, as will other emerging applications of AI to customer service, such as sentiment analysis.

Sentiment analysis: a game changer

Sentiment analysis is also a new and very powerful AI technology that uses natural language processing for customer service interactions. It can analyse spoken and text communication and identify tones of conversations in order to provide an indication of a customer’s feelings towards a company, and its products and services.

Companies can use data from sentiment analysis solutions to monitor customer satisfaction, replacing the post-interaction verbal or email surveys we are all familiar with. Also, sentiment analysis can be applied in real time to raise an alert about a frustrated or angry customer and escalate the interaction. A human agent can take over from a chatbot in these situations.

It’s still very early days for sentiment analysis — it is available only for English language interactions — but it will be a game-changer for the customer service industry. The more data customer service managers have on customer sentiment, the more service improvements they can deliver.

Businesses will be able to compare customer satisfaction across different call centres, with individual agents, and quickly identify products or services that are not being well received. They also will be able to identify chatbots, or even specific chatbot responses, that are poorly received by customers.

Any organisation looking to deliver an optimal customer experience should explore these new AI technologies. At the same time, companies should exercise care in their application. As advanced as they are becoming, they aren’t always up to every aspect of customer care.

It’s been pointed out, for example, that AI-enabled systems for voice-based interactions may not always be able to replicate the nuance, cadences and flows that characterise person-to-person conversations, nor are they usually proficient in injecting humour into conversations. Voicebots, chatbots and digital humans also may not be able to catch onto new jargon right away. Still, embracing these technologies to some extent may enhance your company’s ability to satisfy customers’ support expectations so they are convinced that your brand is worth doing business with.