Data / Machine Learning / Sponsored / Contributed

In God We Trust, All Others Bring Data

23 Jun 2021 9:00am, by

Joann Starke
Joann is a senior product marketing engineer at HPE. She holds a B.S. degree in marketing and computer science, having graduated with honors in the field.

Data’s value continues to increase as organizations look to leverage its value by extracting insights for business efficiencies and competitive advantage. Along with the increase in data value, data-oriented regulations continue to proliferate with greater complexity making availability and reliability of data critical.

Data management has been a core component to achieving insights for data-driven organizations. However, the events of 2020 reminded every organization that data management is the enabler for analytics initiatives because of its ability to ingest, integrate, discover, manage and ensure data quality. The result is a renewed focus and bigger budget assignments for data management initiatives. Yet hurdles remain.

Digital transformation has complicated the ability to achieve trusted insights due to rapid investment in machine learning and AI-driven automation. Edge and multicloud have increased the types and number of data sources from which data needs to be ingested and protected. Then there’s a new generation of workers, known as Gen-D, who come with expectations that all data should be accessible to them no matter where it is located.

Global data access presents a significant hurdle for solutions using traditional data-access models. That’s because these models were designed to deny access, not make it available to everyone. Denial of access means that data scientist and analyst productivity is reduced. According to 451 Research, these workers spend up to 48% of their time requesting access to, or waiting for, data delivery. Clearly, a new approach to data engineering and analytics is required.

DataOps is the new approach being deployed by organizations to accelerate the value of data analytics, data science and machine learning — through a combination of technologies and methods — with the goal of delivering a continuous pipeline of trusted data for applications and users. But integrating data across a plethora of sources remains a manual task and requires data science and developer teams to copy data to another system before they can access it further, which slows down insights.

Enter the modern data fabric, which enables DataOps by automating the processing, transformation, security and orchestration of data across multiple sources. This provides your business with a trusted, real-time view of customer and business data to your applications. A data fabric is defined as “orchestrating disparate data sources intelligently and securely in a self-service manner, leveraging data platforms such as data lakes, data warehouses, NoSQL databases, translytical and others, to deliver a unified, trusted and comprehensive view of customer and business data to support analytic applications and insights.”

Or stated differently, data fabric technology is the key to modernizing data management. Without it, your data science and developer teams spend more time and effort ingesting, integrating, curating and securing data insights.

Modern Data Management

Beyond reinvigorating data management’s value for analytics, the events of 2020 also provided the required spark for the disruption under way for unified data management and analytics. The recently released 2021 Enterprise Management Associates Radar Report reflects this disruption, with 73% of enterprises surveyed looking to move from today’s leading platforms to a more unified data management and analytics approach. The goal of this consolidation is to reduce data engineering efforts across different teams and processes, while also freeing up data scientists to focus on data exploration and modeling.

To achieve these goals, a solution needs support for a wide range of tools, programming languages and APIs, as well as a hybrid approach to data placement. It needs to establish a separation of concerns to allow data teams to focus on how their applications use data, rather than on data logistics, and allow system admins to focus on administration, instead of application-level concerns. Lastly, data management, security and data protection need to be built in.

I am not describing some future vision, but rather a proven technology that is being used across a wide variety of production environments.

HPE Ezmeral Data Fabric unifies data management and analytics into a single platform that spans core, edge and multicloud environments. Tool and language flexibility enables data teams to use tools most comfortable to them, while also allowing legacy and modern apps to access common data sets simultaneously. A global namespace allows direct access to data, files, tables and streams, even if they are located in different clusters and locations. Lastly, built-in platform-level security, data protection and data management are simplified with global policies that let you set and allow the automation and orchestration, to transparently manage daily tasks.

As data continues to grow in both volume and importance, modernize your data management to allow data science teams to deliver trusted insights to your business decisions. Learn more about how to do this by reading this Forrester report.

If you are looking to update your data management strategy, read about the benefits of a modern data fabric and how it can simplify architecture, workflows and life for data scientists and engineers.

Lead image via Pixabay.

A newsletter digest of the week’s most important stories & analyses.