Benson Wu, Co-Founder, CyCarrier

Taiwan-based CyCarrier Technology is an AI startup founded by three top cyber experts namely, Birdman, Benson, and PK. The AI 1.0 platform was released in 2018 and had gained strong traction in various critical infrastructure sectors including government, military, finance, telecommunication, high-tech and etc. Customers were able to achieve optimal defense capability via limited team resources and solved common pain points like too many silo products, shortage of human capital, and ineffectiveness in incident response. At 2019, CyCarrier continued to invest R&D establishing the CyCraft AI Lab, coined from our craftsmanship in AI and future “driverless” cybersecurity operation.

Good guys in cybersecurity always have to deploy layers of fine products and carry heavy workload to address endless threats. In order to flip the table in asymmetric warfare, one must transform linear cyber capability to exponential breakthrough of AI technology – so can hunt the attackers efficiently and effectively.

In recent years, there has been a significant progress in AI technology - machine can even do self-play training and surpass human experts in very short period of time. In 2017, Google’s AlphaGo played 29 million times against itself within 40 days, and in 2018, AlphaGo Zero shorten such training time to 34 hours and championed in Chess, Shogi, and Go. However, rather than worrying about being replaced by AI, humans can aggressively co-work with machines. Security team can off-load tedious investigation works to CyCarrier’s AI-based SOAR (Security Orchestration, Automation and Response) platform, which combines advanced machine-learning algorithms and forensics technology, and boost overall cyber defense capability.
CyCarrier’s AI engine CyCraft has three design concepts: First, sensors and big data enable AI to get the complete situation of digital battlefield; next, the powerful AI analytics platform that simplify and automate complex computation and simulation in a smart single pane of glass; finally, the fusion of global threat intelligence.
Nowadays, enterprises rely on SOC that collected tons artifact-based alerts, while more and more cyber-breaches are leveraging context-based attacks yielding very limited artifacts and hence often very long dwell time. Since these contextual footprints involved several different data types and difficult to be described as rule of thumb, human experts are exhausted in connecting these suspicious dots via SIEM and SOAR.

CyCarrier releasees AI 2.0 in 2019, applying both reinforcement learning and generative adversarial networks (GANs) to self-play normal and malicious simulations – ultimately to achieve “driverless” cyber investigation.

CyCarrier envisions the future cyber operation shall be highly automated and can co-work with experts to boost team capability. The joint defense of human and machine keeps enterprise stay ahead in cyber war.