Skip to main content Skip to footer
Cognizant logo
Case study

The challenge

Witnessing a shift in consumer behavior around contact center engagement, a midsize international airline decided to proactively find new methods to connect with customers who tended to ignore the airline’s Frequently Asked Questions page and directly call the contact center. The airline wanted to reduce direct call expenses while also providing 24x7 support to their passengers.

A particular pain point was the number of customer loyalty program inquiries received. Recognizing the potential for conversational AI to simultaneously reduce call volume and improve customer satisfaction, the airline sought Cognizant’s travel & hospitality technology expertise in developing a customer-facing intelligent assistant to automate these inquiries.

Our approach

Cognizant’s specialized team began by evaluating multiple natural language understanding platforms in the marketplace to determine the appropriate technology stack. After considering the client’s business objectives as well as the platforms’ capabilities, we recommended Google Dialogflow because of its ability to escalate issues to an employee using the LivePerson chat platform. 

We then mapped the conversation flows and handled the chatbot training, which grew in complexity over time as the number of intents grew. To mitigate any intent conflict, the team leveraged a series of best practices built from previous engagements and performed frequent regression testing to ensure the bot had an acceptable accuracy rate. Once that threshold was reached, the intelligent assistant was deployed on the airline’s website. There, it automatically handles both straightforward and complex conversations from customers related to the airline’s loyalty program and escalates issues when necessary.

Automating customer inquiries improves airline customer experience

Cognizant’s Conversational AI team partnered with the airline to deliver a virtual agent that answers over 50 of the most common customer inquiries relating to its loyalty program. Learning over time, the intelligent bot effectively contains and leverages relevant "conversations", reducing the load on contact center agents by responding to all high-frequency inquiries.

1,200 conversations

automated each week by software agent

80%

containment rate by virtual assistant

90%

chat program accuracy