Network Security
Architecture
Deep Instinct is changing the way we look at cybersecurity by harnessing the power of Deep Learning Neural Networks to prevent threats in zero time. The platform’s unique deep learning multi-layer network security architecture is designed for any security need, against any known or unknown threat. It can be applied anywhere in the enterprise, be it network perimeter, endpoints, mobile, servers and VDIs. Effectively protecting the entire enterprise from any type of threat, no matter how advanced they may be.
Hover over a section of the image to find out more
Deep Instinct™ Neural Network
A deep learning neural network is located at the Deep Instinct™ lab. It is the core component of the deep learning network security architecture developed by Deep Instinct™. It continuously learns, reflecting the ever-evolving cyber threat arena. The output of its continuous deep learning process is a lightweight prediction model (D-Brain). The D-Brain is then distributed to all managed D-Clients.
D-Brain (Prediction Model)
The D-Brain is a lightweight prediction model, which is the output of the training phase that detects cyber threats. It is installed on the client software (D-Client). Once installed on a device, the prediction model is used to autonomously detect and prevent cyber threats on the device. This enables on-device zero-day and APT protection.
Deep Instinct™ Cloud Servers
The liaison component between Deep Instinct™ Neural Network and the management console. It sends the latest prediction model D-Brain to the management console, which updates the D-Client.
D-CLOUD
The D-Cloud Intel is the database composed of billions of files, collected from various data sources, and labelled into different verdicts and classes. It serves as the dataset for training and testing the D-Brain. The D-Cloud Live provides a second layer of protection. Using the D-Cloud services, files can be re-classified using the D-Cloud database of intellectual information on known files and the right verdict is updated in real-time.
D-client
The lightweight client software that is installed on the device according to the relevant platform (Windows, macOS, Android, iOS, and ChromeOS) or on the enterprise's network security architecture. It encompasses the essence of Deep Instinct’s™ prediction model (D-Brain) enabling an on-device Deep Static Analysis, Deep Behavioral Analysis, and other key protection layers. The lightweight, autonomous software communicates with the D-Appliance in real-time to receive policy and software updates and to send events.
Management Console
Management and monitoring server, hosted at the organization's data center premises or in the cloud.
