Architecture

Architecture driving decisions and actions

Guavus distributed, edge-analytics architecture is the most flexible available today and makes analytics faster and more decisive. Our unique “analyze-first, store-later” model enables real time responses and automated actions – where and when they are needed.

Guavus Reflex comprises two major components

Big Data Pipeline

First, we have drawn upon 10 years of pioneering big data experience to create a big data pipeline architecture leveraging best-of-breed open source components and integrated them into a hardened platform.  Raw business or network data from disparate sources are ingested as high speed big data streams or batches, then normalized, enriched and curated.  Data is transformed and stored in different forms throughout its journey in the pipeline so that it can be accessed at multiple points along the way.

Guavus Reflex Analytics Engine

The second component is the Guavus Analytics Engine. It performs behavioral modeling, contextualization and reasoning on the various data forms within the pipeline to generate rich insights. The Analytics Engine starts with a comprehensive and extensible set of machine learning algorithms. To these, our team of scientists have added artificial intelligence or “reasoning”. Artificial intelligence enables the Analytics Engine to draw logical inferences even when presented with patterns of events which have not been seen before and which therefore it has not been trained on.

Building on the Platform to Meet Business Needs

Guavus Analytics Application Modules

Utilizing the Guavus Reflex® platform as the brain and base, we have developed purpose-built analytics modules that sit on top of the platform. Each analytics module has a specific function and solves particular business problems.  The analytics modules converse with the platform through APIs, invoking the desired algorithms and pulling the appropriate data to meet the specific needs of the application.  Insights can be viewed through graphical user interfaces for rapid decisions or directly linked to 3rd party systems via APIs for a single pane of glass view or to invoke automated actions.

Customer-Built Application Modules

Guavus Reflex enables platform users to turn their own subject matter expertise into big data, intelligent applications rapidly and easily. Understanding the complexities of integration of all the technologies needed to implement advanced analytics applications is very time consuming and complex, and projects often are abandoned. Reflex is built to eliminate these issues. Guavus has pre-integrated a comprehensive set of technologies and pre-assembled components into higher level services which are easily invoked via APIs, allowing businesses to create their own advanced analytics applications.

Distributed Architecture Makes Analytics Faster and More Decisive

The Guavus Reflex® distributed architecture is the most flexible available today and can be deployed in a distributed or centralized model. Rather than transporting all the data back to a central site to be processed there, Guavus’ platform is based on a distributed architecture for optimal handling of data from separate silos. Regional sites can have smaller implementations of Reflex running at each site, performing the same advanced analytics at edge, in real-time.

By collecting data and analyzing it at the edge, the amount of time needed to trigger closed loop actions or act upon the insights provided is significantly reduced and workflow inefficiencies downstream are removed.  The contextualized data can be then sent back to a central site in regular intervals, aggregating the regional findings and drawing further conclusions, including information from all sites and triggering further automated actions if desired.   Data can be stored in the Guavus platform itself or fed to a central data lake.

Distributed Edge Analytics