As global spending on AI research and development rises, companies are incentivized to roll out AI-driven innovation into new areas of industry and our daily lives.
Taiwanese startup Appier is one of them. With the $80M Series D round last year, it has so far raised a total of $162M from venture capitalists to "innovate in AI for industries beyond digital marketing."
This week, the firm announced in a press conference that a former computer science professor Dr. Shou-De Lin has just come on board as a Chief Machine Learning Scientist. With over 20 years of academic experience in AI, knowledge discovery, and natural language processing, he is the third scientist to be invited to join the team. He called Appier “the first international AI company born in Taiwan.”
During his teaching career at National Taiwan University (NTU), Dr. Lin famously set up the Machine Discovery and Social Network Mining lab and won six championships in the KDD Cup, an annual data mining competition organized by the Association for Computing Machinery.
Also, as a former consultant, he helped Appier optimize a machine learning model that detects hesitant online shoppers, following its acquisition of Japanese startup Emin.
Taking research into the real world
Before joining the team, Dr. Lin said he had met Appier’s Co-founder and CEO Chih-Han Yu in major AI conferences and was pleased to see how the startup tries to keep up with the latest tech trends. “Appier puts a lot of focus on turning research results into new products and services,” Mr. Lin added.
As a former professor and Chief Machine Learning Scientist for Appier, one of his main missions is to bridge the gap and facilitate the synergy between industry and academia.
“I am already immersed with my colleagues and I look forward to both working alongside and guiding them in further developing solutions that not only solve challenges for our customers but also open up new business opportunities,” he said.
Dr. Lin believes the experience in the startup will give him the opportunity to understand and learn from the market that new AI applications can benefit in many ways, too.
He compared building an AI model to taking an exam: “It’s easy to score 60 (out of 100), but difficult to score 90, and the 30 points (for optimization) is what makes all the difference.”
Besides recruiting experts from the university, Co-founder and CEO Mr. Yu also announced in the press conference another way his company promotes the collaboration between industry and academia: by sponsoring the newly-launched Appier AI Chair Professor Program. He said the goal of the program is to “invite experts from around the world to come to Taiwan and train local talent.”
In fact, Appier has been running a program since 2016 to sponsor selected students to attend top conferences like NeuroIPS, ICML, and ICLR and present their research findings. So far, over a hundred have benefited from the program and stood on the global stage to showcase Taiwan’s research capabilities in AI, according to the firm.
With past experience in the university, Mr. Yu and his team understand well how important yet challenging it is to turn insights derived from research studies into valuable, real-life applications. “Researchers has a quite different perspective than us,” he said. That’s why the company strives to build a bridge between the two worlds by offering travel grants for students and inviting them to share what they’ve learnt with the team.
Trends and challenges
"The evolution of AI will bring about a paradigm shift in technological development,” said Dr. Ming-Syan Chen, Executive Vice President of NTU. “Exploring innovation in AI is one of the most important topics in the global technology community.”
He pointed out three areas where AI has the most potential: healthcare, manufacturing, and the service sector. The technology not only change industries but “form its own industry.”
Mr. Yu said more efforts should be made in several areas to unleash the potential of AI, including ensuring data integrity and accuracy and preparing the public for the wide application of AI technology.
While the technology enables a number of new services, Dr. Chen said, it’s vital for companies like Appier to acquire industry-specific domain knowledge to build what users need.
In response to this, Mr. Yu shared his startup story: “In the early days, when we just started Appier, a lot of our solutions failed because we didn’t know what users need,” Mr. Yu said. “Now, we learn from them to decide what solution we can provide to help them.” This is what he called “market-driven innovation” based on a user-centered approach.
The next step, for Appier, is to see if they can adopt such solution across different industries and markets.