Request a demo We can arrange a demo to show you the capabilities of our products

Our products are used to:

_reduce the uncertainty of Machine Learning models,

_prioritize enterprise data efforts,

_support experts in the ML loop,

_improve the quality of ML models, especially in multi-class settings with complex ontologies,

_reduce data footprint and compactify ML models so as to be used by the Internet of Things applications,

_improve gaming experience via more challenging and realistic AI in games,

_create intelligent advisory systems from pre-compiled building blocks.

LITL for Cybersecurity

Human experts are both the gem and the bottleneck of cybersecurity operations. While there are plenty of tools for gathering and analyzing the ever-increasing amounts of data, it is the human expert, who has the ability to select and interpret the relevant information, taking into account the situational context. 

Yet, SOC experts’ decision-making process is difficult to codify. Their tasks are complex, and they are often highly specialized with respect to attack types. How to delegate that?

According to Deloitte and CyberRes, in the Next-Gen SOCs humans and machines work together to manage cyber risks. This is the direction CyberOps is headed.

Label in the Loop (LITL) technology from QED Software enables creation of Artificial Experts: software agents that perceive the environment in which they are situated (including SIEM data, human expert activities, etc.) and react to the events occurring in that environment (e.g. notify the human experts, surface examples of similar incidents from the past, etc.). Their decisions depend on the predictions of ML models trained on data collected from observing the human experts at work.

  • On 10% of the data, we achieve over 80% of the quality of the model trained on 100% of the data (better models faster and cheaper).
  • The method of extracting expert knowledge reaches over 110% of the quality of your best expert.
  • Our cold-start mitigation methods (selecting the initial data batch) achieve a stable result and the quality of the 80th percentile (or higher) of the quality distribution of the (unstable) baseline method.

Let’s set up a meeting!