TouchPal’s next-generation keyboard app at TechCrunch Hangzhou

2 min read

Globalization has become a mainstream trend among Chinese tech companies over the past year. But if we look back, this trend could be dated back years ago when China’s first group of tool apps found traction overseas. Shanghai-based keyboard app TouchPal is among the Chinese startups that pioneered this trend.

As a third-party keyboard service, TouchPal now claims over 600 million users and a predominating 95% of them are outside China. While the keyboard still the company’s core product, it is now gradually developing other portfolio apps for fitness and content, but their focus is still on the overseas market.

Keyboard app may not sound a sexy business, but it’s a very basic service for every user. “Ten years ago when I left my job at Microsoft and co-founded my new company, we thought that keyboard is not that sexy but it could be a good starting point before we move on to cool projects like Facebook and Google,” TouchPal co-founder Michael Wang said at TechCrunch Hangzhou.

But as the team keeps going, they found that keyboard app turns out to be a very successful business and it is also strategically important. “The more we developed innovative keyboard features, the more we believe it’s related to AI, natural language processing and how the machine can understand and see the world,” Wang added.

TouchPal has developed lots of interesting features from mistyping corrections to slide and gesture input and Michel believes the introduction of new technologies is opening up more potentials of the sector. “In the late development stage, we thought we almost have tapped into all of the new visions of the keyboard, but we were wrong and got to find lots of interesting areas. For example, we have released an AI assistant for keyboard, which is embedded in keyboard and helps users to forecast the weathers, recommend restaurants, etc.”

TouchPal’s user distribution is quite global, not only in the US, Europe but also most part of Asia. The company tries to empower data collected from the users by leveraging deep learning technology.

“We found that the user behaviors in different countries and areas are quite different. Learning those languages through machines, we found something in common. People in certain areas are discussing certain topics more often. For example, World Cup is so popular now. To some extent, we can even forecast the result of political elections. Google catches that statistics from search and TouchPal from the keyboard,” Michel shared.

Of course, data collection would raise privacy concerns among users. “We are very serious about user’s data privacy. A certain amount of language data would be needed to train the language model. If users don’t want to give their data they can disable the option. Similarly, Google and Microsoft all have their corresponding data collection policies and agreements. In addition, we also spend lots of resources in the compliance with GDPR (General Data Protection Regulation),” he said.