Data scientists need to harness and analyze Big data to gain insights that are business critical.
Many times, they need to aggregate Big data with core business applications’ data.
With the pace of change in any business, end users need to have on-demand access to data spread across the enterprise.
In addition, some of the end users have become ‘Citizen Data Scientists’ and have a need to rapidly build the analytics views to aid business leaders
for timely decision-making. To effectively address the needs of power-users, end users, data scientists, and citizen data scientists,
there is a clear need for a bi-modal modern A / BI platform.
Key tenets for Modern A / BI platform are summarized in diagram shown.
Mode 1 is for the governed BI deployment for all users where the BI artifacts are developed and managed by power-user or IT developers.
While Mode 2 is for rapid and agile A / BI approach for citizen data scientists, end users, and data scientists. In both modes of deployment,
security and access control are paramount for any business. Underlying applications’ security rules are expected to be enforced in A / BI artifacts.
Many A / BI products approach the problem of aggregating data from across the enterprise by first developing an intermediate data store
(a Data Mart or a Data Warehouse) that stores a copy of the data from source systems. With this approach, the security and access control
rules of the underlying source systems are by-passed.
Additionally, data ETL routines must be developed, maintained, and deployed to make this
approach work. When the business conditions or requirements for analytics artifacts change or when the source systems’ versions change,
the entire infrastructure of ETL and intermediate data store must be upgraded. At times, this approach can result in data discrepancies between
source system data and intermediate store data.
Resolution of these data discrepancies can take a lot of effort and may impact the
perception of end users that they are not dealing with trust-worthy data.
ETL routines are typically run overnight and therefore,
the analytics is not real-time or near real-time. For certain business decisions, this ‘stale-data’ can be a major problem.
The shown diagram depicts the traditional approach and summarizes its challenges. This entire approach is laborious, expensive, and slow!
eQ’s approach is to keep data where it belongs – in the source systems. eQube®-BI can be deployed in a bi- modal manner.
It interactively mashes-up data from multiple systems with stunning visualizations to reveal the ‘story’ behind the data for Actionable Insight.
It honors and leverages the underlying applications’ security and access control rules. Therefore, end users see only the information they are
authorized to see.
Analytics artifacts can be consumed by end users in an eQube®-BI defined portal or as part of a SharePoint portal or as part of
a web-based 'For-Purpose' App or on any mobile device. It has in-built powerful scheduler that can publish the analytics artifacts for end users
in both Mode 1 and Mode 2 deployment approaches. Diagram shown depicts the eQube®-BI approach for A / BI.
eQube®-BI has in-built event management system (EMS) that generates A / BI artifacts upon certain events in underlying system (s),
such as upon executing certain step (s) in a workflow or upon state change of an object or a database record. In addition, end users can generate A / BI
artifacts in real-time on-demand.
Sentence-based analytics is fully incorporated in eQube®-BI. End users can type in their questions in plain English
in a search bar and generate A / BI artifacts due to the interaction between eQube®-BI’s powerful natural language processing (NLP) engine and its data
virtualization layer with semantic capabilities.
In addition, eQube®-BI efficiently deals with streaming /sensory data as well as Big data stores and Data Lakes to provide aggregated view across these data sources and core business systems (such as: PLM, ERP, MRO, Supply Chain, Asset management, Logistics, ALM, etc.) for critical insight. The underlying architecture is enterprise-class scalable architecture with highly optimized in-memory cubes that scales out to support thousands of end users.
eQube® believes in keeping data where it belongs – in the source systems.
However, if a customer has already invested in a Data Warehouse or Data Mart (s) or Data Lake, eQube®-BI can easily leverage it as a source for
A / BI capabilities. In addition, it can mash-up data from the Data Warehouse or Data Mart (s) or Data Lake along-with the data from other enterprise
systems (legacy, COTS, files, etc.)