Menu Content

Economy

Skelter Labs, an AI Firm Developing Machine Intelligence Technology

#Power of Businesses l 2019-03-25

Business Watch

© Skelter Labs

Now it’s time to take a look at a Korean business bringing about changes in the global economy with some new ideas. 


The company we’re going to introduce today is Skelter Labs, a promising artificial intelligence firm that develops machine intelligence technology. Let’s hear from chief product officer Choi Won-joon. 


Artificial intelligence or AI is one of today’s hottest topics, as computing power and data-based applications and services are becoming increasingly important. 

Indeed, there has been remarkable development in AI-related technologies in recent years. Most workers at Skelter Labs company are AI specialists who have accumulated expertise in the industry. Our CEO and myself, for example, previously worked at Google, whose AI program, AlphaGo, stunned the world in 2016 when it defeated one of the world’s best Go players in a 5-round match.


Skelter Labs has a number of AI experts with various professional backgrounds at established companies such as Google, Samsung and the Korea Advanced Institute of Science and Technology, commonly known as KAIST. 


Company CEO Ted Cho, in particular, is a leader of Korea’s first venture startup generation. Cho studied computer engineering at Seoul National University and at KAIST. He founded a company called Saerom Technology in 1993 that made unique software designed to send a fax through a computer, not through a fax machine. In 1999, he developed Dialpad, the world’s first free-of-charge internet call service. Afterwards, he served as the R&D chief at Google Korea. In 2015, he established Skelter Labs to develop core AI technologies. Here again is Mr. Choi. 


We focus on developing applicable AI in voice recognition, speech synthesis, conversation, vision and context recognition. Speech synthesis technology helps imitate voices of a specific person in a natural way, while conversational AI can understand the context of dialogue and even comes up with proper answers. Vision technology aims to read and analyze images, classify products or detect flaws. Context recognition involves identifying people’s real-time activities and their personal preferences.


The goal of Skelter Labs is to develop machine intelligence technology designed to understand users’ daily lives and help them make better decisions to improve their quality of life. 

To this end, the company focuses on the development of conversational AI applicable to chatbots or AI speakers, a context recognition engine that perceives the users’ situation through data measured by various devices, and vision and speech technologies based on deep learning. 


Skelter Labs’ context recognition technology, in particular, is a step above the rest.


© Skelter Labs

Most AI-powered conversation engines give just one answer to a question. If you ask such an engine to find sneakers, for example, it simply shows a pair of sneakers. But our product follows the flow of conversation, has situational awareness and offers an answer that fits the current context. So, if one wants to find sneakers, our engine understands the conversation and incorporates factors such as one’s gender or brand preference before making a final answer. 


The conversation engine Skelter Labs is currently working on is capable of recognizing a user’s overall situation and making tailored recommendations. For example, when reserving a hotel room, most AI-based conversation engines process information such as a preferred date and room type sequentially. But Skelter Labs’ engine can process all inputs simultaneously. For instance, a command such as “Make a reservation for a hotel suite with an ocean view in Jeju Island on the 15th day of next month,” would pose little challenge for the firm’s technology. 


Skelter Labs’ innovation opens the possibility for a sophisticated AI assistant that advises users in a host of different capacities, such as when to take an umbrella when it looks like rain even before the user asks about the weather. 


The company’s technology and potential have gained due recognition, as it secured nearly 10 million U.S. dollars in investor backing in 2017. The company has used these funds to spur its technological development even further, applying its AI technology to the manufacturing industry.


Our vision technology is used at a local car factory to identify defects of components. At present, even the most advanced and highly automated factories still rely on human employees to detect defects. But humans may make mistakes, due to fatigue, poor eyesight, or other conditions. Our solution aims to help manufacturers improve their defect detection rate considerably as well as their overall working environment.


Quality control is one of the last phases of production. But this highly-important step can be difficult, costly and time-consuming. Skelter Labs’ AI solution has addressed these problems with machines that can identify defective products more efficiently than humans. Drawing a positive response among Korea’s manufacturers, the technology will be introduced in Japan next month. 


We’re considering exporting our defect detection system to Japan, which has a highly-advanced manufacturing sector. At present, Japan’s economy looks to be in relatively good shape despite ongoing challenges such as an aging population that is leading to drastic decreases in the working population. The manufacturing sector remains strong, and Japan is particularly known for component manufacturing that requires complicated procedures and thus a high-demand for skilled employees. We believe our AI solutions will prove popular and competitive in the Japanese market


Skelter Labs hopes to export its AI solutions overseas, starting with Japan. The company will then move toward the global AI market. As the tech firm develops more advanced technologies and AI is increasingly understood as a way to improve people’s lives, there’s little doubt Skelter Labs’ solutions will be sought after and widely utilized.

Editor's Pick