Needed: More Automation, More Low Code
There’s too much data and demand for insights on a widespread basis to leave it to manual processes. “The sheer volume of data growth, vast complexity of multi-jurisdictional usage laws, increased speed of business, and need to apply more predictive and prescriptive intelligence are posing significant challenges for enterprises and pushing executives and their data managers toward a critical tipping point,” said Wills. “That inflection point is automating the examination/monitoring of metadata (what is known about the data) and data itself (analytics, usage, location, access, privacy) to trigger both automated and human interaction. Some would call this robotic process automation for the data world.”
“With rapidly changing business environments, data continues to play an outsized role in operations,” said Jim Sears, vice president of professional services at Boomi. “Yet to ensure real-time access to data, automating the flow of data as well as any updates to it or associated privileges has lifted a burden off of data teams. These days, everyone needs to do more with less, and automation and low code have become the answer.”
At the same time, “the past year has highlighted some cracks in automation, such as how easily automations can break, which is happening more often than ever because of constant software updates,” Sears cautioned. “Expect to see more companies extending data automation to include intelligent automation, using AI to intuitively understand and implement automations across applications and processes.”
Business and technology leaders “are under pressure to find new ways to disrupt their industries by launching new data services fast,” Sears added. “One solution they’ve turned to is low-code technology with drop-and-drag capabilities that allow for efficient updates. Low-code tools give time back to leaders to help them and their teams more quickly spin up applications, make data connections, and focus on larger business challenges at hand.”
Cutting Through the Silos
Data silos have plagued enterprises since the creation of the first database. “Data is also becoming more omnipresent. It’s everywhere, from our watches to the applications we use in the cloud,” said Jerod Johnson, senior technology evangelist at CData Software. “However, fundamental issues with data remain. It is still siloed and difficult to access, and with more data than ever before, it has become a major pain point for employees to work with the data they need to drive their business forward. The persistent problem we’re seeing today is not a lack of data; it’s the accessibility of it.”
Business and technology executives understand this accessibility issue: “They hear it from their business units all the time,” Johnson said. “What keeps them up at night is finding a way for the data to be usable by everyone across their organization, while still keeping their data ecosystem secure and governed. IT leadership will find relief from their woes when they can stop gatekeeping data and instead provide business units with secure, self service data access. Modern data connectivity solutions give even nontechnical users the ability to build bridges between their various SaaS applications, data repositories, and platforms—in the cloud or on-premises.”
One of the biggest problems for enterprises “will be data integration,” agreed Karanjot Jaswal, co-founder and CTO of Cinchy. “Enterprises have more data than ever before; it’s more fragmented—with multiple sources, formats, and silos—and it’s become harder to find and extract than ever before. The complexity of data management creates a tremendous amount of work, trying to simply gain access to data that already existed in the organization and to be able to understand what other data is out there, to then be able to further bring that information together to actually deliver net new capabilities for the business.”
Expect to see more movement toward the “liberation,” or freeing, of data in the year ahead, Jaswal added. “More than just new implementation, this fresh mindset around data liberation will help bring relief. We all tout data-enabled collaboration, but through it all, our data foundation is app-centric, not data-centric. Moving forward, we need to separate data from the applications used to create and store it. This helps bring the data together in a unified network. It’s a change that’s both simple and powerful.”
Data Keeps Sprawling
“Data sprawl and governance have become more difficult as data becomes increasingly distributed across the data center, edge, hybrid, and public cloud infrastructure,” said Radhika Krishnan, chief product officer at Hitachi Vantara. “More data is expected to shift from centralized analytics to localized analytics. Data is exploding quickly. Today’s average enterprise environment is said to include more than eight data lakes. Although there is more valuable data than ever, it is more difficult to manage than ever.”
Data integration and preparation “are critical as the data comes in from multiple sources,” Krishnan added. “The notion that the public cloud is for agile, cloud-native workloads and data centers are for traditional workloads is no longer the reality. Companies are looking for a simple and consistent hybrid cloud operating model that has the cloud-like agility and automation and can seamlessly connect their applications, data, and infrastructure across on-premises, near-cloud, and public cloud.”
Multi-cloud, on-premises, and edge environments “will become increasingly common to meet evolving demands,” said Krishnan. “This approach also comes with tremendous complexity as customers need to manage and integrate with multiple different cloud operating models across applications, data, and infrastructure. With data accumulating faster than IT teams can keep up with, it’s important that businesses consider ways of addressing the costly resources it takes to maintain the vast volumes of storage.” A DevOps approach also will help address demands for more modern data application environments, he added.
Needed: More Responsive Infrastructures
“There is a growing understanding that the ‘application-first’ and the consequent ‘integration-centric’ approach to traditional enterprise data management are reaching its limits,” warned Jeremy Bentley, head of strategy at MarkLogic. “As more application data sources are integrated to feed digital acceleration initiatives, the enterprise data infrastructure becomes more and more complex—and very brittle.”
This situation “is preventing data systems from being able to respond readily to change and encourages the deployment of additional data silos, adding further complexity,” Bentley continued. “This Catch- 22 frustration is what business leaders want removed, as successful business is all about rapid innovation and being able to respond nimbly to change.” Bentley advocates a “data-first” approach. “This is why there has been increasing interest and investment in data mesh, data fabric, and semantic knowledge graph architectures.”
Aligning With Customer Experience
Modernization is key, but it relies on “transformation exercises such as CX transformation, database rationalization, data lakes, and lakehouses, as well as customer interactions and policies,” said Steve Zisk, senior product marketing manager for Redpoint Global. “In particular, companies that significantly improved CX for digital and mixed-mode customer interactions experienced much higher growth than business-as-usual organizations.”
Talent Shortfalls
The market for data scientists and engineers, information security, data storytellers, and interpreters “is growing much faster than the number of graduates entering these fields,” Zink pointed out. “In some cases, cross-training and enabling data-savvy business users will help. Company culture and data-driven plans can also help organizations compete for and retain the best talent.”
Many issues dominate and will continue to dominate the agendas of enterprise data teams, but the goal needs to be helping their business reach and understand their customers. Prepare for another exciting year.