Cyberupt focuses on the use of model-based machine learning and probabilistic fusion to build intelligence systems for a variety of domains and applications. Our systems are able to ingest and enrich data at high volume and high velocity, sourced from a variety of sensors and sources. This also includes data from the cyber domain and transactional systems. Both supervised and unsupervised machine learning are used to enable the discovery of threats and anomalies, even for rarely-occurring events.


The type of sensor and data source is largely determined by the domain of operation. Input data can range from financial transactions, network packets in the cyber domain, to data from optical and RF sensors, tracks from radar and other MTI sensors, OSINT and HUMINT.


A Cyberupt system is deployed on a robust distributed network of Datacenters, employing a Service Oriented Architecture. The use of Cloud technology ensures that data is available to users where and when needed. Where the customer has an existing infrastructure this is exploited as far as possible and augmented where required.


We also work with system integrators and product owners to embed the Cyberupt building blocks into existing implementations to add machine-learning and information fusion capabilities.