Blogs

IBM launches new Integrated Analytics System with Machine Learning

Post Comment
Principal Analyst and President, Pund-IT, Inc.

Information analytics has never been a “one size fits all” proposition. That applies to the hardware and software technologies  organizations employ, the information being parsed and the goals of specific projects. So it’s worth examining how individual vendors approach analytics and the way they evolve their solutions and services to reflect changes in commercial markets.

That’s certainly the case for IBM’s Integrated Analytics System, a new solution that builds on previous company database, software and hardware innovations, along with its longstanding support for open source projects and organizations. Let’s take a closer look at that.

IBM’s data analytics history

It should be noted that IBM has been deeply involved in data analytics and business intelligence for decades. The company’s work in hierarchical databases resulted in the SABRE system American Airlines deployed in the 1960s. In 1970 an IBM researcher, E. F. Codd, was the first person to propose the concept of relational databases. The company’s work in the 1970s on its System R database resulted in the creation of SEQUEL (Structured English Query Language, or SQL) for manipulating and retrieving data.

IBM introduced its Db2 database in 1983, and since then has stayed at the forefront of information management and analytics development. Its acquisition of Netezza and its data warehouse technologies in 2009 reflected a growing shift toward leveraging open source and industry standard technologies to benefit analytics efforts. More broadly, since 2005 IBM has invested over $26B in advanced analytics, including the development of its PureData System for Analytics and support for efforts like the Apache Hadoop and Spark projects.

Are you Data Science Experienced?

The company is also at the forefront of critical analytics and data science performance issues: creating solutions that allow processes to be managed more efficiently and performed where or near where information resides. In fact, IBM is branding the new Integrated Analytics Systems with: Do Data Science Faster.

Why are these points important? First, data science depends on rigorous processes that can be undermined by inefficient collaboration or subtle complexities. So easing or eliminating those issues can help deliver more accurate and valuable results.

In addition, analytics solutions that require data to be imported from and results to be exported to a host system impose substantial time and cost penalties. That can impact the value of whatever insights are gleaned, and limit how results are used by data scientists, business analysts and their employers.

In contrast, IBM offerings address those issues and deliver resulting benefits. For example, the company’s Data Science Experience focuses on providing flexible collaboration services for data scientists, so they can easily share their models, findings and expertise, and work together more effectively. In addition, the Data Science Experience also supports model building, deployment and scoring functions for a wide variety of projects and use cases.

IBM’s solutions also enable analytics processes to be performed on both incoming data and data at rest (including historic information) with results delivered at or near real time. That’s crucial for customers, including banks, financial entities and credit card companies that need quick alerts to fraud and other criminal activities. In addition, access to speedy predictive analytics can enhance services, from loan applications and approvals to weather and traffic updates to tailoring online advertising according to customer preferences.

The IBM Data Science Experience also provides a host of online resources that enable data scientists and other interested parties to easily access, work with and find support for the company’s advanced technologies and services.

IBM’s new Integrated Analytics System

So what is the company delivering in this newest solution? In essence, the IBM Integrated Analytics System builds on and considerably extends the company’s PureData System for Analytics and the Netezza data warehouse and analytics offerings.

Like those solutions, the new System leverages the common SQL engine and industry standard tools, and supports numerous data types and platforms, including IBM Big SQL, Apache Spark, Db2 Warehouse on Cloud and Apache Hadoop (including Hortonworks Data Platform). In fact, the new offering is the first to come after the expanded partnership IBM and Hortonworks announced in June.

In addition, the Integrated Analytics System comes with a variety of built-in data science tools which are designed to help organizations get up to speed with advanced projects. For example, customers can use the embedded Data Science Experience and Apache Spark capabilities to develop machine learning-based business automation capabilities. Since the new solution leverages IBM’s common SQL engine, it can also easily utilize public/private clouds to support workloads and business processes.

Finally, the Integrated Analytics System incorporates hybrid transactional and analytical processing (HTAP) which supports predictive analytics, transactional and historical data on the same database at accelerated response times. IBM also plans to add support for HTAP to the IBM Db2 Analytics Accelerator, which will enable the system to transparently integrate with IBM Z mainframe infrastructures.

Why is this important? Because the addition of Integrated Analytics System capabilities will enable Db2 for z/OS on Z and Db2 Analytics Accelerator to provide simultaneous high-performance transactional and analytical processing. That will be welcome news for many of IBM’s core Z mainframe infrastructure and Db2 for z/OS clients.

Final analysis

What’s the bottom line on IBM’s new Integrated Analytics System? Two things stand out. First, the new solution underscores the company’s continuing investments and leading position in advanced analytics. Those include its own homegrown innovations and support for/partnerships with open source organizations, including Hortonworks and the Apache Foundation.

Second, the Integrated Analytics System reflects how IBM tailors its efforts for the needs of its customers. Like every other business, those organizations are looking for a competitive edge that can help them maintain or grow market share. But they also need proven, dependable solutions and services to get there, not science experiments cooked up on a whim.

The new Integrated Analytics System is designed to help customers achieve those goals fully and successfully. In other words, Do Data Science Faster isn’t simply a marketing tagline but a reflection of IBM’s deep understanding of advanced analytics and its ability to help customers address their fundamental technological and business needs. 

 

IBM would like to thank Charles for this contribution to the Big Data and Analytics Hub and we encourage readers who are interested in the IBM Integrated Analytics System to learn more here. This piece is part of a collection of reports, which can be found at http://www.pund-it.com/  

© 2017 Pund-IT, Inc. All rights reserved. 

About Pund-IT, Inc. Pund-IT™ (www.pund-it.com) emphasizes understanding technology and product evolution and interpreting the effects these changes will have on business customers and the greater IT marketplace. Though Pund-IT provides consulting and other services to technology vendors, the opinions expressed in this commentary are those of the author alone.