Lablup, a startup that offers big-data and machine learning solutions, has received 2 million won in investment from K Cube Ventures and Stonebridge Venture Capital.
Lablup offers a PaaS (Platform as a Service) to automate the process for cloud services, server maintenance and scaling for machine-learning environments that generally require a lot of system resources. As the development of machine-learning technology progresses, and engineering environments become more complex, the company says there is a need for more scalable and standardized computing platforms.
The startup was originally founded in 2015 by Jeongkyu Shin, formally the lead developer on once-popular blogging platform Textcube. With 10 years’ experience in various ventures, Shin has a Ph.D in physics and has also worked as an Adjunct Professor at Hanyang University.
Lablup’s flagship platform is named Sorna, a distributed code execution engine designed for computational research and the training of deep-learning models. The system is flexible and can be scaled from a single machine to several clustered servers using its ‘Docker’ technology.
Sorna is offered as a pay-as-you-go cloud service, and is also available as an open-source platform for those that want to deploy using their own hardware. To try out the system for yourself, you can visit ‘CodeOnWeb’, a website which teaches programming languages commonly used in the STEM industry. The site was developed in-house and uses the Sorna engine to compile code.
Lablup is now developing a streaming service which uses a machine-learning model, as well a machine-learning based natural language processing API. The company also plans to offer the open source version of Sorna as a solution to organizations and companies with server farms, and its cloud version to companies which don’t have enough IT infrastructure in place to support projects.
In addition to these plans, Lablup is also looking at developing an AI framework based on Sorna’s technology.
“As innovations made by machine learning and data-based systems increase the range of applications in IT environments, limits to the scalability and convenience of systems are inevitably being reached,” says CEO Shin. “Through our solution, we will support engineers by offering a ubiquitous experience in complex computing environments.”
