SuperCharger Startup Highlight: Interview with Ken Yeung, Co-Founder of Clare.AI

By Natalie Wong, Program Associate of SuperCharger

Ken Yeung, Co-Founder & CEO, Clare.AI

Ken Chi Kwong Yeung is the Co-Founder and CEO of Clare.AI, a white-label artificial intelligence chatbot to help financial institutions provide 24/7 personalized customer experience across platforms.

Previously to building Clare.AI, Ken held various roles in the Investment Banking field developing award-winning global multi-asset trading platforms using machine learning and NLP.

Clare.AI is one of the 8 companies which was admitted to the SuperCharger FinTech Accelerator Program 2.0.

Q: How can AI chatbots help Hong Kong banks and financial institutions engage with their customers?

I think the AI Chatbot has many use cases on how it can help banking. Essentially the biggest part we see is in customer service. Customer service in Bank is painful now. You probably have experience. You have to call the bank all the time, you have to wait a long line, and usually there is phone banking but you have a long list of options (to go through), and it’s very hard to follow each options. It’s a very painful problem. And there is not many online channels now where you can reach your customer.

Q: What makes Clare.AI stand out from other bank/budgeting chatbots that already exist in Europe and the US?

I think one of the biggest difference is that we focus on customer support, as we see compared with others like in Europe and the US.

I think what have been seen in Europe and the US is that, they are more focused in personal financial management linking to your bank account, like checking your spending, budget, and so on. Maybe that’s really related to the culture there. Budget spending is important as a student as they only have certain amount every month, they have to keep track of that.

Versus in Hong Kong or in Asia, the spending habit is different, the problem is different as well. And right now Clare AI is more focused on in Hong Kong and also in Asia, with more specific problem we have in this region, and that’s customer support: reaching the bank, how to get efficient of the bank, both from the bank side and from the customer side.

Banks have been spending so much money in customer support. A typical bank spends around 10 million USD just on the customer support but people are not happy, customers are not happy, and you are not happy as well. So that’s why we’re tackling this problem.

Q: Would you also like to tell us about your Cantonese chatbot?

Yes. Right now we are focusing on Cantonese, cause that’s a major language in Hong Kong.

We use a proprietary algorithm where we gather data — specifically in Cantonese — public data, proprietary data, and we build a set of grocery, dictionary, and intends. Intends are basically like the intentions of each questions, so we can figure out what you are asking.

For that, the advantage is that we can easily plug into the banking system, so they don’t have to retrain the questions. If we plug in the module of Cantonese, we can capture what people want to say in a sentence.

Q: So people can use exactly the same functions with Clare.AI in Cantonese as they could in English?

Yes, they can just say naturally what they say in Cantonese in Clare AI. Cantonese is a very special language. You have slangs, you have mixed English & Cantonese, and that’s really difficult compared with English.

Some companies tried to do a translation from Cantonese to English, but it won’t work, if you tried Google Translate (you know what I’m talking about), it gives you very wrong translations. So that’s why we’re building Cantonese capability and that’s how we got our clients because no one else is doing that.

Q: Some say that Artificial Intelligence might replace human jobs in finance and accounting. What is your take on that?

I think that’s probably the negative side of it. I think there’s actually more of a positive side. It’s not replacing. It’s actually assisting the customer rep.

As I said there’s a long waiting time, people are not happy. And we’re just trying to assist the customer representative to answer repetitive calls. 80% to 90% of the calls are repetitive, which means that people are answering the same questions all the time. But we don’t want people to just answer all the time, doing a boring job. In this industry actually 25% of the people will leave the company, the turnover is very high.

So we’re just trying to assist. Instead of answering repetitive calls, they will handle complicated calls. Secondly, they will train the AI bots, so actually we see more new challenging job opportunities. These people who are used to answering same questions, they can move to more high value jobs.

Q: What is Clare.AI’s plan in 2017?

Right now we are focusing on building the product.

We already have the customer pipeline, so we need an engineering team, and in order to do that we need funding. We are actually at the stage where we have 1 paid pilot and 10 other prospects. So we are looking for a fund just to focus on building the pilot, as well as to be capable to building the ongoing customer, to be ready on that. That will be our main focus this year.

To get in touch with Clare.AI, please email

Email the writer: