The art of conversation

CASE STUDY

Improving efficiency through virtual agents

The big picture In today’s world, customers expect an ever-greater depth of experience and value from the services they seek out. However, overcoming the obstacles to delivering a superior customer experience has never been a greater task than it is now. To do so means investing in a digital future, one in which automation has been deployed for organizational efficiency, and technology and systems have been upgraded to serve customers in better ways. Importantly, updating policy systems, automating processes and integrating with partners will be the keys to unlocking the next wave of productivity.

The challenge Our client, a leading general insurance firm, faced a number of these obstacles. The overall customer experience and the productivity of their agents was affected by inefficiencies throughout the company’s contact center.

What our client needed

Our client needed a solution that could improve their contact center’s efficiency, reducing the wasted hours of agent time and the wait times for customers. Ultimately, this problem required an automated process that could identify the intent of callers and provide a voice response to the simpler queries, freeing up agents to tackle the complex queries that needed their attention. Our design and technology expertise made us an ideal partner in creating better digital value for the client’s 20m customers and making the most of their 19bn data points.

The contact center – which accounted for over 20 thousand hours of agent time each day and nearly 35 million conversations a year – was being slowed by a large number of requests and limited agent time. Agent time was being wasted handling queries that were sent to the wrong department, leading to longer wait times for customers. The company’s security authentication process also took a significant amount of time to complete.

The client needed automation to unlock the next level of efficiency in their business.

A transformative solution We worked with the client to categorize the various intents behind the customer calls, identifying sixteen categories that would help model for automated responses. The next stage of the process involved using a large dataset of nearly 12,200 recorded conversations to train and test the model that we developed.

After taking these steps, the model was designed using machine learning and NLP neural network-based techniques, before undergoing several iterations to achieve greater accuracy.

We boiled down customer intents down to 9 categories to model for automated responses

A transformative solution

What we provided

Leveraging GCP, we deployed a voice bot that could identify and handle the simpler queries of customers. We were able to create a conversational design flow for three of the simple intent categories, including policy updates, cancellations, and document requests, which altogether were driving 35% of total call volume.

Natural language understanding

Leveraging conversational design frameworks:

Intent & entity recognition

1. Opening comments

“What’s the weather like tomorrow?”

2. Acknowledge intent & clarify process

Natural dialogue management

3. Sense of closure

“It will be sunny and 20° C”

Voice Platform

Your backend database

Action resolver*

Natural language generator

Not only did this have the effect of freeing up agent time, but it also created an easier, more streamlined process for the handling of customer calls.

The results We found that nearly 78% of customer conversations could be grouped within 4 of the simple intent categories, meaning that the automated voice bot could effectively manage a large number of requests. Rolling out the solution resulted in an estimated 40% of time saved for the client’s customer care agents. The immediate impact

Our solution will boost the efficiency of the contact center over time, with four intents making potentially good candidates for automation, each having average 95% rate of accuracy. These intents also formed nearly 40% of conversations within the contact center, indicating more room for scale through conversational ai solutions. Long-term benefits

Four of the call intents we defined are candidates for automation with 95% accuracy

Estimated 40% of time saved for customer care agents

Enable better decisions with Fractal Fractal is one of the most prominent players in the Artificial Intelligence space. Fractal’s mission is to power every human decision in the enterprise and bring AI, engineering, and design to help the world’s most admired Fortune 500 ® companies. Fractal product companies include Qure.ai, Crux Intelligence, Theremin.ai, Eugenie.ai & Samya.ai. Fractal has more than 2,300 employees across 16 global locations, including United States, UK, Ukraine, India, and Australia. Fractal has consistently been rated as India’s best company to work for, by The Great Place to Work ® Institute, a ‘Leader’ by Forrester Research in its Wave ™ on Specialized Insights Services, Computer Vision & Customer Analytics and as an “ Honorable Vendor” in 2021 Magic Quadrant ™ for data & analytics by Gartner.

Corporate Headquarters Suite 76J, One World Trade Center, New York, NY 10007

Get in touch

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7

Made with FlippingBook - PDF hosting