Web Agents with Natural Language Processing identify and capture all relevant publicly available information for a Cyber Security event.
Much of the publicly available information is no longer available, either because the affected organisations have removed documentation or the publishers have deindexed the content. Our approach hunts out this content and once found makes its own local copy so that we have a growing library of reference material against which to perform our analysis. This library is insulated from the fragility of the publicly available information.
Machine Learning algorithms detect and extract metadata from this unstructured information.
The extraction of metadata from these events allows us to draw patterns and understand themes which may not be immediately obvious when looking on a case-by-case basis. For instance, did you know that the majority of employee-involved hacks have happened over a weekend (employees target and are targeted when offices are 'thin-on-the-ground') and yet companies still reduce their on-site security staff over these periods? It is this type of insight that metadata extraction provides us.
Artificial Intelligence is combined with human analysis by industry-leading individuals to produce detailed, high quality case studies for each event.
The combination of AI and human analysis gives us the best of both worlds. The AI surfaces insight that traditional human analysis may have missed, for instance the common use of an attack vector or encryption approach, whilst the human analysis provides us with key insights into the cause and effects of the event and allows us to build up a invaluable store of lessons learnt.