Under the airline’s conversational AI roadmap, Cathay is optimising its digital customer experience even further with the most cutting-edge language AI technologies. The new conversational AI developed in partnership with Fano Labs will improve the efficiency and productivity of the airline’s AI chatbots, enabling them to provide more accurate responses to customers’ queries.
As demand for travel returns, Cathay is anticipating an increase in the volume of call and chat traffic, with post-Covid travel enquiries already growing rapidly in the past few months. The airline has also seen a shift in customer behaviours, turning from traditional hotlines to digital channels such as WhatsApp and WeChat for more instant responses. Digital channel usage has experienced monthly growth of 10-20% over recent months, as Cathay continues to add back more flight capacity.
To build an AI chatbot that provides a superior digital experience for customers, the most important yet challenging task is to find large volumes of high-quality data to train the chatbot so that it can understand what people are asking, and why, to provide proper responses. Such chatbot training is often time consuming and expensive, requiring human “AI Trainers” to look at historical data to see how customers interacted with the chatbot, manually extract, clean, and annotate useful portions of this historical data, and then feed it back into the chatbot for learning.
To address these challenges, Cathay and Fano Labs have become one of the first in the market to have successfully developed and launched the Conversational AI Training Excellence platform that uses AI technologies to train AI chatbots. The “Trainer AI” leverages Fano Labs’ Callinter product to analyse interactions with customers, including voice calls, live chats, WeChat messages, and WhatsApp messages, and automatically group these data into different categories for further processing, before presenting the results to a human supervisor for review and approval. The AI chatbot then learns from the data extracted by the trainer AI.
With this system in place, Cathay expects to reduce the time needed to train chatbots by 50%, meaning a regular 4-6 week learning cycle will be shortened to 1-2 weeks only. The solution enables Cathay to take on a significantly greater volume of training data and ramp up its ability to discover new topics and improve the accuracy of chatbot responses.
Cathay Director Digital and IT Lawrence Fong said: “As a company with more than seven decades in Hong Kong, we are committed to supporting local tech startups such as Fano Labs and innovating together to build our digital leadership – a pillar for achieving our vision of becoming one of the world’s greatest service brands. This collaborative effort in Conversational AI technology is the latest example of how we can work together with startups to develop mutually beneficial partnerships and ultimately bring benefits and enhanced service to our customers.”
Fano Labs Co-Founder and CEO Dr Miles Wen said: “Fano Labs, a homegrown AI company, was founded with a mission to bring the best multilingual language AI technologies to our clients to generate business value. Our partnership with Cathay started in 2019 under the introduction of Hong Kong Science and Technology Parks Corporation. Throughout the past few years, we are honoured to be entrusted by Cathay to take part in executing the company’s conversational AI strategy. This successful collaboration is a strong showcase of Hong Kong’s thriving innovation and technology ecosystem.”