Date
13 November 2018
Fano Labs uses speech recognition and natural languages processing to help out at clients’ call centers. Photo: Bloomberg
Fano Labs uses speech recognition and natural languages processing to help out at clients’ call centers. Photo: Bloomberg

Startup Fano Labs uses AI to understand Chinese dialects

Horizons Ventures, billionaire Li Ka-shing’s private investment arm, has announced an investment in artificial intelligence (AI) startup Fano Labs, the firm’s first bet on a Hong Kong-based company.

Fano Labs chief executive and co-founder Miles Wen recently sat down with the Hong Kong Economic Journal to discuss how the company uses AI technology to understand Chinese dialects.

Here are excerpts from that interview:

HKEJ: Natural Language Processing (NLP) has been a white-hot technology that many tech companies jump into, yet, analyzing Chinese dialects is not a popular area. Why did you choose to focus on this?

A: There are many companies developing their NPL technology capabilities in major languages, such as Mandarin Chinese, English, Spanish and French. But very few companies work on dialects, which is indeed an important capability for enterprises. Say, a Hong Kong-based corporate would need to handle calls in Cantonese, and for a Sichuan-based company, 70 percent of the communication is in Sichuanese.

The market demand is huge. Yet, we need a vast amount of accumulated data for training the AI model, which is considerably less for dialects, so are the specialists who know the specific dialects.

So we would try to work with large-sized corporate clients, using their database to improve our AI model, which would then offer services to these corporate clients.

Q: How does the funding from Horizons Ventures help grow Fano Labs’ business?

A: We already have a considerable number of corporate clients, and we are backed by Hong Kong’s Innovation and Technology Commission, as well as three angel investors back in 2016; so we didn’t have an urgent need for raising capital at the time.

But Horizons Ventures is a great firm. It has a heritage of investing in deep-tech companies, and it previously invested in speech and NLP technologies such as Siri, Viv, so we agreed to accept that funding, for the future of Fano Labs.

Q: Can you explain the business model of Fano Labs?

A: Fano Labs uses speech recognition and natural languages processing to help out at clients’ call centers. We process and analyze the call center’s interactions with our NLP system, to help them to evaluate staff, check on customer feedback and train new joiners. We charge our clients an annual fee for software licensing.

We are targeting over 200 corporations in Hong Kong which have their own call centers, as well as tens of thousands of companies in mainland China. According to our estimate, the potential market size will reach billions of dollars.

Q: With so many peers in the market, how do you step up against the competition?

A: AI technology companies are mainly competing for talent. With our close ties with the University of Hong Kong (Fano Labs is a spin-off from the University of Hong Kong), we have an advantage in recruiting fresh talents.

Q: What do you think about the so-called China-US battle for AI technology?

A: The success of the tech industry depends on talent, not location. China’s AI technology is fast catching up with that of the US, while the US tech industry has accumulated more experience than any other country worldwide.

One interesting thing about the AI sector is the people in the industry usually have doctorate degrees. And you will find that many of them, in fact, graduated from universities in the US, even taught by the same group of professors.

Lots of Chinese scientists who graduated from US universities are now going back to China to join the AI industry, but what if they leave the country sometime in the future? It would totally upend the whole China AI sector for sure.

This article appeared in the Hong Kong Economic Journal on Jan 19

Translation by Ben Ng

[Chinese version 中文版]

– Contact us at [email protected]

BN/CG

Hong Kong Economic Journal

EJI Weekly Newsletter

Please click here to unsubscribe