AI Chatbots: A Must-Have for Customer Service

Everyday AI At Work Blog Series

When they first emerged chatbots were simple – mostly just web browser pop-ups that could only send straightforward messages like “Hi, how are you?”  Now with the integration of latest technologies they can be a key customer service element in web based customer interactions.

Hang Seng Bank announced in April 2017 their plan to launch the “Virtual Assistant” service by the end of 2017. The chatbot prototype, powered by IBM Watson, is offered on the website and mobile channels for internal testing currently.


Hang Seng Bank announces their plan to use chatbots for customer service. From left to right: Barry Chan,  Partner and Banking, Financial Services and Insurance Industry Leader, IBM Global Business Services, Paul To, Assistant General Manager, Head of Digital Banking, Retail Banking and Wealth Management, Hang Seng Bank, Mimi Poon, Director, Banking and Financial Markets, IBM

Taiwanese E.Sun Bank (玉山銀行) launched AI chatbot (玉山小i) in April 2017 to provide a personal financial e-advisor service for its customers. The service is based on IBM Watson and was the result of E.Sun Bank’s partnership with IBM and instant messaging app ‘LINE’.

Many of the world’s leading companies are deploying AI platforms, this includes financial institutions and banks, and they are seeing that these deployments are helping to improve customer interactions, reduce backlogs and evolving into facilitating a wider range of self-service offerings. The deployments are also improving business process accuracy, and leading to reduced costs and improved efficiency.

AI Makes Bots Customizable

Integrating the ability to understand and process natural language and understand the tone and sentiment of the conversation has greatly increased the scope and complexity of what can be delivered by using an AI enabled chatbot. This has encouraged business users to explore business use cases and options that deliver additional value to both the consumer and the enterprise.

For example, one company might use a chatbot to help customers with simple online product orders. Another could build a bot that assists professionals with research and communicates using natural language.

In addition, companies are increasing the data available for research by AI, in refining its interactions, so that the extent of the interactive service can further reduce the requirement for human interaction or a manual offline process.

Man and Machine, Working Together

Many companies are deploying virtual assistants to help their customer service agents provide more personalized guidance when interacting with a client. This is leading to better customer satisfaction as in the example of Bradesco, one of Brazil’s largest banks, deploying a virtual customer service solution that allowed its employees to help customers get answers to questions faster. After a few months of training the Watson empowered service can answer questions with 80 percent more accuracy, leading to more satisfied customers. As the AI technology gets more knowledgeable with every interaction, the Bradesco deployed chatbot continues to improve the customer experience and service solution.

Building Emotional Intelligence into Bots

Cognitive computing capabilities can derive information about users’ intentions through the words they use, thereby allowing AI to add emotional intelligence to the chatbot. The goal of every conversation with a chatbot should be to ensure the user feels understood, especially in situations that may be sensitive.

Many companies are already using AI technologies that can detect user tone. For example 1-800-Flowers, the floral gift retailer, launched GWYN, (Gifts When You Need) an AI-powered gift concierge. GWYN interacts with the online customer using natural language. It asks the user questions about the occasion and sentiment for their purchase, then offers a tailored gift suggestion. This process demonstrates that GWYN understands the human intention behind the purchase, rather than just acting as an automated Q&A tool.


1-800-FLOWERS.COM taps AI to manage tens of thousands of customer interactions occurring daily to deliver memorable moments (Credit: 1-800-FLOWERS.COM)

As AI progresses, chatbots will become more skilled in natural communication, emotional intelligence and data analysis. And they will increase in their value to any interaction between man and machine, leading to better outcomes and smarter decisions.

By Mimi Poon, Director, Banking and Financial Markets, IBM China/Hong Kong Limited

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s