Using AI to Better Direct Customers

June, 2019

Conversational AI/ AI | Chatbot | Customer Experience | Data Mining
By asking open questions, a robot can identify the reason for the call and then either handle the call itself or transfer the call to a live agent. Excitement surrounding Artificial Intelligence (AI) has been building steadily for some years.
Excitement surrounding Artificial Intelligence (AI) has been building steadily for some years. Thanks to increases in computing performance, the ability for a computer to access large volumes of data and take complex, probability-based, human-like decisions has increased. That, in turn, has resulted in many practical applications that use AI-techniques being developed.

Improving interactions

In the contact centre there are tasks that AI can perform that can improve interactions. Consider what a contact centre represents to an organisation and its clients. From the organisation’s perspective the contact center is their way of funnelling all interactions with customers through a single point to ensure nothing is missed. However, that funnel can have an unwanted side-effect of putting all customers in the same situation, regardless of their history and importance to the organisation.

For example; customers calling in to a contact centre may find that they always wait in the same queue, are then passed around between agents, having to re-explain their issues each time, before they reach the person they need to speak to.

The AI workflow

Using an AI ‘bot’ to have an initial conversation with a client is a great way to enhance the customer’s experience. A Natural Language Processing (NLP) engine can absorb the many different combinations of words that a client might use and determine their intent. By asking open questions, a robot can identify the reason for the call and then either handle the call itself or transfer the call to a live agent.

Information collected from the customer during the bot conversation can be compared against similar data collected from other customers to make an AI judgement on which is the most suitable agent to transfer the customer to. AI might determine that a customer from a particular region of the country should talk to an agent from the same region as this creates a natural bond between agent and customer, which serves to increase customer satisfaction.

Configuring a bot and training it to recognise the phrases customers may use will not be done in isolation. Many actions may need to be taken to control a session and give the customer the best possible experience. Such tasks include database lookups to determine which agent is best for a customer, re-queueing a session to a different queue, updating a database to tell the system to contact the customer at a later time. All of these are also part of the flow of a session. Contact Center configuration tools should give supervisors a single interface through which they can perform these tasks. By making contact center system management tools easy to use, AI is more likely to be used more frequently and more effectively, which results in a better experience for end customers.

For more information about Sytel’s products, just talk to us.