September 23, 2022

4 min

Intelligent Virtual Agents Enhance the Customer Experience

AI Brings Efficiency and Richer Interaction with Cloud Contact Centers.

Image of a woman working on a desktop computer.

Contact centers have evolved into customer engagement centers, managing how companies’ interface with their customers. How a company can efficiently and effectively create a better customer experience is also evolving. Leveraging artificial intelligence to accomplish tasks and help a customer self-serve is at the forefront of all discussion as of late. So, let’s dive into how you can make your customer experience better using AI and automation.

Artificial Intelligence

As AI has advanced, the engagement center has incorporated agents into customer interactions, and It is important to think of these as virtual agents. They are treated like people; however, they do not need benefits, sleep, food or water, and they work for your organization 24/7/365. 

Most companies price the Intelligent Virtual Agent (IVA) just like agents, meaning you pay per agent, and you figure out how many simultaneous agents you need for your customers. These IVA’s can be used to do intelligent call steering, troubleshooting using FAQs or true self-service. 

There are several vendors out there that have abstracted the backend and given you a low-code, no-code environment. This allows you to pick and choose your AI backend. For the most part, there are three major players in AI backends: Google, Amazon and Microsoft. And all three offer similar services: natural language understanding, natural language processing, text to speech and analytics.

Workflow Automation

Workflow automation is used to tie systems together. It creates a way of getting the data into the AI so that it can do things more efficiently and provide rich responses to the customer. These systems are crucial to the effectiveness of the intelligence of your virtual agents. This also allows your agent to do things like update records in your CRM on what the IVA did with the customer and what steps were taken. It also addresses fulfillment tasks, based on what transpired while communicating with the IVA. Having the intelligence to complete automation steps throughout the process brings in needed information about the customer’s journey in interacting with our non-human agents.

AI Tuning and NLP/NLU

There is some constant care and feeding required for these virtual agents. That usually involves matching what is said to what the agent can match those phrases to. In the AI world, those are referred to as utterances. AI technology is constantly self-learning, but it requires training phrases. For instance, if I were calling into a healthcare provider and said, “I need to see a doctor,” that phrase is completely different than, “I want to schedule an appointment.”  Having training phrases for the various ways in which I can ask for a service is crucial. The more an AI is used, the more it can figure out things and the smarter it becomes, but it can be a daunting task to figure out the tuning.

CDW has been involved in many customers’ tuning exercises, and we have developed a component of our managed services tilted continuous tuning. This gives the NLU the ability to understand, and this service gives the AI a more lifelike and conversational aspect for the customer interaction.

Tying It All Together

Intelligent Virtual Assistants can operate the same no matter what the incoming channel might be. If I start with a chatbot virtual assistant, I can reuse what work I have done on the backend on a voice channel. I can create a voice IVR to use natural language understanding to get me quicker to the option in an IVR, allowing me to skip to the exact option I want, saving me time in the IVR.  

I can use these virtual assistants to fulfill tasks such as password resets, giving an account balance, rescheduling a flight or booking a vaccination appointment. They can stay in the conversation and transcribe the call allowing for automated ways of putting notes into a CRM of what happened with the customer. 

Workflow automations can tie various tasks and systems together, allowing for greater self-service fulfillments. All in all, we can service the customer no matter the time or day, even on holidays. These agents can handle 100 percent of the call or merely augment the physical agent, allowing them to serve the customer better and more effectively. This gives the customer a better way of interacting with us and a faster way of getting the information they need or the tasks they need done.

Story by Scott Petersen, a Contact Center Technical Architect for CDW’s Services Research and Development team. Scott has experience in data center, contact center, wide area network and switching technologies. He is a graduate of the Lean Sigma Six Greenbelt program, as well as an AWS Certified Cloud Practitioner, Amazon Connect, Webex Contact Center Expert, Webex Experience Manager, Five9 Certified, Inference IVA Certified, and Nice inContact(RingCentral) certified.