Unify: Architecting the Missing Link in Data Management
Data users are facing a disconnect.
On one end, we have sophisticated tools that can help make data analytics a visual experience, with enough features to turn numbers-minded business users into citizen data scientists. On the other, we are now connected to a lot more data.
Yet, companies are struggling to turn those huge data stores into insights. It is where the biggest disconnect lies.
“Every organization yearns for consistent, well-governed data that is easy for the business to access and use, as well as simple for IT to deliver and manage. But you are right, organizations are struggling to get there,” says Robert Eve, senior data management strategist and the thought leadership and digital content executive, TIBCO Software.
Solving the disconnect is urgent. On the demand side, business users are clamoring for “access to all data, everywhere, at any and all times,” as they navigate the relentless headwinds from COVID-19 measures in addition to all their other data-driven business initiatives. On the supply end, data is experiencing phenomenal growth, “doubling every 2.5 years.”
For Eve, the disconnect comes down to how companies store data within IT. It “doesn’t always match how it is used by the business, thus lots of transformations are required.”
Real-time data is an additional complexity. “It is a nascent domain in many IT teams, but its importance is soaring as real-time responses are critical when engaging customers and optimizing operations,” he adds.
Plus, stricter data security measures and new government regulations that scrutinize the handling of data at rest and in transit makes data management that much harder.
No room for data fantasies
Analysts understand the rising problem. They say the answer lies in embracing distributed data rather than focusing on centralizing it.
“Gartner calls this a logical data warehouse. Forrester uses the term data fabric. Data-as-a-services is another frequently used term,” says Eve.
No matter what label or acronym the industry attributes, it comes down to a simple truth that you need a dose of reality before tackling data management. “All recognize the fact that it is impossible for organizations to physically centralize all their data. Instead, data virtualization lets organizations provide one “virtual” place to go for data consumers to access data and IT to provide it,” says Eve.
Next, companies need to have a strategy to tool up for “next-generation data management.” “Gartner’s advice to consolidate with their data management tooling in vendor suites such as TIBCO Unify that combine metadata management, master data management, reference data management, data catalog, data governance, and data virtualization within one integrated solution,” says Eve.
Data management should not be an IT problem alone. Businesses can chip in by increasing their citizen data engineering pool and offering business domain advice. “Work together to assess your needs and skills. Then be smart about maximizing the value each side can contribute, for example, IT using TIBCO Data Virtualization to provision hundreds of reusable data services that the business can quickly mix and match to address their changing needs,” says Eve.
Speed is essential — especially in today’s dynamic business landscape where pockets of opportunities can add relief to stressed cash flows. “With data at the core of digital transformation, victory goes to the swift. Adopt agile methods such as data virtualization to avoid the long SDLC associated with ETL and data warehousing,” Eve advises.
4 ways to tackling the disconnect
Companies should stop hoping for a single data management answer. Instead, Eve proposes that they should take a pragmatic approach by using a solution that takes a multi-pronged approach to eliminating the disconnect. TIBCO calls it Unify and tackles the disconnect in four ways.
First Unify focuses on simplifying data complexity. Eve calls it radical simplification of complex processes. Here, data virtualization will be key.
“It is a lot easier to integrate data using data virtualization than it is when trying to make physical integration with all the mappings, ETLs, staging schemas and destination schemas work. And rather than using six or more different tools to manage metadata, master data, reference data, transaction data and streaming data, with TIBCO you can use one integrated suite, Unify,” Eve adds.
But where TIBCO deviates from other competitors is its drive to enable citizen data engineers. While they may not be full-fledged data scientists, they can ease their workload. “Business domain experts can update master and reference data directly, for example at Panera Bread, the store managers maintain store configurations in the master data repository,” says Eve.
Third, Unify embeds artificial intelligence and machine learning to take over manual data management processes. These include metadata discovery, data quality, model design, code generation, query optimization, resource allocation, etc. “This added intelligence not only saves data management resources, it increases business agility and optimizes compute and storage resource usage,” Eve says.
Finally, Unify is not alone. A “rich catalog of specialist partners” puts expert knowledge and experience in companies’ hands.
Bridging the gap
The data management disconnect is a result of companies looking at data science from a single dimension. They wanted insights and sought these by centralizing their data stores. It is fine when these insights are not crucial for daily decision making, finding strong correlations between current and historic data, identifying outliers, and exploring future opportunities.
COVID-19 compressed the timeline and made these insights more urgent. The tools were there, and the data was available, but the data management infrastructure was not ready. Replacing their current one was also not an option in today’s ultra-lean times. Eve, however, quotes Plato, “Necessity is the mother of invention,” in highlighting the hidden opportunities.
“When we put together TIBCO Unify, we recognized that that ‘rip and replace’ wasn’t a path to success. Instead we let you leverage existing data assets, silos, integrations, and more. This means you can start anywhere and focus on the highest return on your data management investments first,” he describes.
To succeed, Eve advises companies to take baby steps by delivering PoCs that deliver business value while helping raise competency levels. To manage data well, both work in tandem.
“For example, although one place to go for consistent, secure, well-governed data is your ultimate goal, the best way to get there is via a series of smaller scale data virtualization efforts. Rather than trying to address all your master and reference data in a big bang effort, with TIBCO you can address a few selected domains. The same is true for metadata management and the data catalog,” he says.
With the right experience and expertise, you can then look to bridge that nagging disconnect between data and insights across the company.
“With these successes under your belt, you are better equipped to execute broader enterprise challenges such as standardizing your enterprise data model, expanding to support more lines of business and users, and scaling to achieve stricter SLAs,” Eve adds.
Photo credit: iStockphoto/wildpixel
