overviewadaptersapplicationscore toolsetscognitive operating systemsanomaly detection


Intellective ai Technology Overview

Figure 1: Technology Comparison

Intellective AI has built the Foundation Platform for processing video, SCADA, cyber, image, and analytics data using artificial intelligence. The heart of the foundation is the Cognitive Operating System. The operating system utilizes a neuro-linguistic cognitive engine as a basis for Cognitive Analytics. The FOUNDATION includes FOUNDATION Video, FOUNDATION SCADA, FOUNDATION Image, FOUNDATION Cyber and a multi-sensor fusion foundation, such as the Homogeneous FOUNDATION (video-video) or the Heterogeneous FOUNDATION (video-SCADA-network). Intellective AI’s FOUNDATION Platform is capable of automatically discriminating between normal and abnormal events from a variety of input data. It learns and understands compound collections of heterogeneous multi-input group data without the need for supervised learning. It does not rely on predefined patterns to identify events and anomalies but instead learns patterns and events by observing the incoming input data and generating lexical information about what it observes. As illustrated in Figure 1, the FOUNDATION Platform excels at identifying the unknown and the unexpected. The Unknown-Unknown is often the most alarming and costly risk.

foundation platform overview

Figure 2: FOUNDATION Platform Overview Diagram

An overview of the FOUNDATION Platform is shown in Figure 2. The system is a highly scalable, multi-input machine learning system that uses GPU based artificial neural network technology. It is the first Cognitive AI System to automatically create actionable intelligence and alerts. It observes, learns, and responds in real time, identifying threats and problems as they happen. The FOUNDATION Platform observes, learns, and adapts to incoming data streams without bias or knowledge a-priori. It trains itself with unsupervised learning through observation to establish a running baseline model of normal events.

Figure 3: Observe, Learn, Analyze, Rank, and Report

As shown on Figure 3, the system adapts to environmental changes by continuously learning and adjusting the model. As it learns what is normal, it creates a hierarchical set of abstract memories that describe the associated events corresponding to normal activity. It then uses these memories to evaluate future events to determine whether they fit within the normal range or are abnormal and need special attention. As these abnormal events are detected in real time, the system issues alerts that identify the concerns.


The core platform, as shown in Figure 4, provides user interfaces, reporting, and data management services. It can be deployed as a standalone system or in cloud-based environments. The Cognitive Operating System can be augmented with modular plug-in application extensions to enable capabilities for a broad range of market sectors and industries that can benefit from neural network powered artificial intelligence input and data analysis. Examples of such applications are Video, SCADA, Cyber, Image, Forensic Analytics, Financial, and Medical. Intellective AI’s business model is to license the Cognitive Operating System Platform to allow industry partners to build on and develop Cognitive Operating System extension modules. To facilitate third party usage of Cognitive Operating System extensions, SDK and interface libraries are being developed for various plugin types.

Figure 4: Intellective AI Foundation Architecture Diagram