Artificial Intelligence

3 steps to infusing AI in the contact center

By Jafar Adibi

0 min read

Crawl-walk-run, introducing three specific ways to improve your contact center

1. Crawl: How can we enhance IVR performance?

To enhance the IVR performance of your contact center, Natural Language Processing (NLP) analyzes hundreds of millions of calls to determine why people are calling and how all the sentences coming from the customer-agent interactions can go into one meaning category.

As the example shows, there are many ways to ask the same thing. A robust, comprehensive and sophisticated NLP program, powered by AI, is key to understanding why a customer is calling and route them accordingly.

An IVR without Natural Language Processing will struggle to identify why your customers are calling. Enhanced IVR performance can help you and your contact center by:

  • Discovering customer intent: Know why people are calling you to improve your entire operation and to have more and better informed contact center agents.
  • Identifying the best IVR setup (e.g. new IVR exits): Sometimes just adding a new exit or reshuffling position two and three is all it takes. Those couple of seconds in every call can bring a significant improvement at the end of the year.
  • Finding agents in need of coaching: Identify which agent performs better for each topic. A three-month analysis of a contact center can reveal that some agents have better performance in certain topics. This comprehensive data allows you to map those agents to certain topics and look at KPIs such as Average Handle Time (AHT) and estimate the Customer Satisfaction (CSAT) performance, offering routing optimization suggestions that can improve the entire interaction.
  • Providing training recommendation: An AI-powered IVR can offer training recommendations by identifying which agents are in need of additional coaching.

2. Walk: Enhancing agent’s power with an agent assist

Artificial Intelligence can help you enhance agent intelligence and overall responsiveness. Before every call, AI can look at the entire knowledge base, the CRM and every conversation transcript. AI assistance analyses these information sources and, when a person calls in real-time, the voice is transcribed into text. Agent Assist can then browse the knowledge base, the whole conversation and all the available data to fetch the most relevant document on the topic and present it to the agent.

Agents come and go. They do not stay for long. That is the reality of the business and agent training is a hassle. Can AI streamline the agent training process? Can we bring knowledge under the agent’s fingertips in a snap? Though the agent might not have any clue about the customer’s questions, what if AI can deliver the right information that could help the agent solve the issue in 10 seconds after the conversation begins? An agent assist can help you with:

  • Real-time actions and quick shortcuts: Reduce the need to leave the customer on hold or having to transfer the call to a different agent or to a supervisor.
  • Smart notes: Summarize the entire conversation and show the transcript for the agent to accept it or modify it.
  • Automatic data entry: When someone calls, an agent assist can make the right application pop-up with the name and address of the customer so the agent is not delayed by having to type in any data.
  • Escalation assistance: It is very painful for the customer when their call is transferred and they have to describe the entire situation all over again. With a powerful agent assist, the complete customer record is summarized, extracted and delivered in just a couple of bullet points to the next agent or supervisor.
  • Real-time analytics and error detection: the agent assist can monitor the call to make sure the agent goes through the standard protocol and follows the established script. The information extracted from this monitoring can later be used to offer training recommendations that can happen in real time or after the call.

Crawling through an NLP-assisted IVR is how you know why your customers are calling. The sooner you know this and start improving the way your contact center carries business, the sooner you will get on your feet with AI.

Then comes the first baby steps with an agent assist, where you will experience the empowerment that such a strong AI tool can give to your agents.

Now you are ready to run with AI. Join me for the next blog as I will run by your side to the final step — Run: Improving customer experience with a virtual agent and learn about all the benefits of having a 24/7 virtual assistant ready to help your customers.

What Is AI and How to Transform the Contact Center in 3 Steps


What Is AI and How to Transform the Contact Center in 3 Steps

Download to learn more about:

  1. What is (real) artificial intelligence?
  2. How AI is transforming the contact center
  3. Three steps to infusing AI in the contact center


Jafar Adibi

As Head of AI and Data Science, Jafar leads AI for the Talkdesk enterprise contact center platform. Prior to joining Talkdesk, Jafar developed advanced Machine Learning solutions for understanding people behavior and consumer intelligence at PwC, Reunify and Incentica. Jafar holds a Ph.D. in Computer Science from the University of Southern California and is a recipient of multiple national and international awards.