Qraft

Founded in 2016 and led by Marcus Kim, founder & CEO, QRAFT Technologies has been on a mission to transform investing with artificial intelligence. QRAFT was started by a group of people passionate about quantitative investing, who found success trading with their own money in their spare time. However, they quickly realized the limitations of traditional quantitative investing, and grew frustrated at both the amount of work and short life span of the strategies they were using. 

QRAFT’s investment experts’ partner with our teams of data scientists, researchers, and data engineers to apply AI technology to data processing, investment research, stock selection, portfolio construction, and risk management. With the advancement of Big Data and our own cutting-edge technology, QRAFT’s employees are on a mission to transform investing with artificial intelligence.

Leadership

Marcus Hyung-Sik Kim

Marcus Hyung-Sik Kim is the esteemed founder and CEO of Qraft Technologies, a pioneering leader in the realm of artificial intelligence-driven investment solutions. A distinguished alumnus of Seoul National University, Marcus holds a bachelor's degree in electrical engineering. During the course of his graduate studies, he developed algorithmic trading program for stock market, demonstrating his exceptional aptitude for both finance and technology. Collaborating with his engineering colleagues, Marcus has been a strong advocate for the application of quantitative and AI powered investment models, which ultimately led to the establishment of Qraft Technologies in 2016.

Process/Philosophy

Kirin API is QRAFT’s proprietary data processing API which pre-processes all data to build a bias-free environment. Third-party data is filtered through Kirin API before being used in the Alpha Platform. The Kirin API is a fully automated engine capable of receiving financial market’s structured data from a variety of sources, and subsequently preparing these data sets to build a simulation environment free of learning biases (survivorship bias, forward-looking bias etc.). Several ways in which raw financial data is prepared for AI model training include security meta data mapping, consideration of fundamental data release date, and consideration of macroeconomic data release date.