CybelAngel Included In European AI Startup Landscape

CybelAngel is proud to announce our inclusion in the European AI Startup Landscape by Hub France AI. CybelAngel was selected from more than 150 applicants as a “Trusted AI Start-Up”. This is a notable recognition of the leading-edge work by our engineering and product team on our Machine Learning capabilities.  CybelAngel has several advantages going for itself in machine learning.

The first is that CybelAngel ML models have seven years’ worth of training data. Beginning in 2015, CybelAngel conducted its first mass scan of the internet. Today we collect three billion new data points daily. This gives the CybelAngel Data Science team a massive historical data set to train our models on. With more than 100 terabytes’ worth of training metadata, our models are highly advanced at predicting digital risks. If you want to hear from the experts themselves, watch Marie Sacksick, CybelAngel Senior Data Scientist, discuss their models here

Active learning strategy

Training data alone doesn’t explain our success using machine learning – another critical aspect is our Active Learning Strategy. There is a continuous feedback loop between data scientists and cybersecurity analysts. As analysts evaluate the criticality of incidents, they report back to data scientists the criteria they used, which items were of high value, and why. This in effect helps our machine learning models to mimic the business smarts and cyber expertise of our analysts.  Read more about our models here:

ML models

The third ingredient in CybelAngel’s Machine Learning success is specialized ML models. Each ML model is specialized, some having only a couple of features while others include more than a thousand. Thanks to constant optimization, the daily Alert feed reviewed by a CybelAngel Analyst has been divided by half over the last 3 years. If you want to learn more about our scoring models read our blog post, Innovative Machine Learning Model for Data Leak Detection.   CybelAngel has invested in machine learning to power our digital risk protection solutions. We’re exceptionally proud to have this success recognized.