5 Data and AI Trends for 2019

What are some of the big data-oriented trends for enterprise programs and culture in 2019? Here's a look.

Jessica Davis, Senior Editor

January 7, 2019

6 Min Read
Image (kathayut) - stock.adobe.com

New year, fresh calendar page. If 2019 looks anything like 2018, you can bet that data, analytics, machine learning, and other forms of artificial intelligence will be part of a developing plan in enterprise IT organizations. Plans will turn into pilots. Pilots will turn into production.

But the year ahead will not be without challenges.

For instance, 2018 may very well be remembered as the year where the data breach or data leak went a little too far for many consumers. Facebook revealed several of these events, and its CEO appeared before government committees, answering questions about privacy and security. That wasn't last year's only challenge.

So what's ahead in 2019? InformationWeek has interviewed and corresponded with a number of industry thought leaders and asked them what to expect. Here are some of the trends they identified for the year ahead.

Data and AIOps

It's been a decade since the industry started using the term and concept of DevOps, which is defined in many different ways by many different people, but comes down to a philosophy of cooperation between software development and operations to create value for the company. The terminology also encompasses a faster and more scalable development process and the concept of infrastructure as code.

Today we are seeing other functional technology areas being joined together with "Ops" to create a faster, more scalable development iterations designed for business value. In 2019, look for DataOps and AIOps to be part of the conversation. DataOps appeared on the Gartner Hype Cycle for data management for the first time in 2018 as an "innovation trigger."

"DataOps will become much more important over time," said Dan Potter, VP of product management at Attunity, an Israel-based data integration and big data management solutions provider.  "How do I take those same DevOps principles and apply them to data to deliver on business objectives." 


Maybe your organization is already openly multi-cloud, but some organizations, such as Capital One do choose a preferred provider for cloud services.

There are reasons for standardizing on a single provider. It's just easier and simpler to work with a single company, both in terms of internal operations and in terms of standardizing your technologies. But there are arguments for using multiple providers as well. While it's simpler to use a single provider, everyone wants to avoid the dreaded vendor lock-in -- when it gets too difficult to migrate your software infrastructure to an alternative provider. Maybe your preferred vendor has changed its terms of service or pricing or quality-of-service guarantees. It's good to keep your options open.

You may already have a multi-cloud strategy. Even if you claim that you are all-in with one provider, you may not be, according to Dave Russell, VP of enterprise strategy at Veeam, and who previously worked as a distinguished analyst at Gartner.  

"Some organizations say they are not multi-cloud today, but they probably are," Russell said. "They may not be aware of it, but somebody in their organization has probably deployed a different SaaS application." Those rogue implementations are more likely to be drawn under the purview of IT in 2019. Russell believes that this year we will see data center people embrace multi-cloud as a reality. 

Workforce, skills, and hiring

The job market for those with specialized skills -- for instance, data science, machine learning, or Python -- has been very tight and will continue to be tight. Organizations are poaching talent from competitors and from universities. There's more demand than supply of these skills, and organizations have been evolving to adapt with more effort to create self-service tools that a business generalist is able to use.

"It's very difficult to hire very specialized people," Russell said. In some cases, organizations may be coming up with new workforce strategies to meet this challenge. For instance, some may be reframing how they think about it. Instead of hiring specialists, Russell said organizations could think horizontally and hire generalists with a range of skills and the ability to learn. These workers should understand the strategy and be in it to deliver a business outcome.

Another approach some companies have employed is to create small teams, with each member bringing a different skill. Such a team could include a statistician, a software developer,  and a business expert, for example. 

Data ethics

Algorithms and machine learning had been pitched as a way to eliminate human bias, but it turns out that when we train our algorithms with the data that we have -- which includes that bias -- the outcomes will be biased, too. The issue has been around for a while and has drawn greater attention over the past few years. Now it's at the point where it has become a top agenda item for some of the thought leaders in big data and data science. For instance, Doug Cutting, known as the father of Hadoop, and chief architect at Cloudera has been spreading this message recently. Late last year he told InformationWeek that business and government need to pay more attention to protecting data privacy and to ensuring data is employed in a way that promotes fairness rather than perpetuating racial, gender, or other kinds of bias.

Look for more industry and thought leaders to address these issues in the months ahead.

Data security and privacy

Data security and privacy were big issues last year, and they are likely to continue as a top concern in the new year. That's something that enterprises will have to watch as they move ahead on their own plans to operationalize data analytics, machine learning, and other artificial intelligence.

Last year at this time, many organizations were scrambling to get ready for GDPR (the European Union's General Data Protection Regulation), which now regulates business behaviors regarding consumer data and privacy. GDPR went into effect in May 2018, and some organizations were still unready. California put its own consumer data privacy rules into effect in 2018, too.

The world is still waiting to see if enforcement of these regulations will have any teeth, but the first enforcement notice was issued against Canadian company AggregateIQ Data Services Ltd. If that name sounds familiar, it's probably because it's one of the companies associated with some of the Facebook data privacy scandals of 2018.

So as we enter 2019, there hasn't been a lot of action in terms of enforcement, but that could change.

"While to date, there haven't been any fines levied against US companies for breach of GDPR, we don't expect this to be the case for long," said Laura Sallstrom, head of global public policy at consultancy Access Partnership. "We expect more frequent enforcement and fines."

On the federal level in the US, in spite of plenty of hearings with social media company executives that addressed data privacy, there hasn't been any serious action to create new laws or rules to regulate it. Sallstrom noted that may change in 2019.

"It's likely that a Democratic-led House will reinvigorate a push for a federal data privacy regime," she said. "But it's unlikely that such a big piece of legislation will have an easy path through congress."

About the Author(s)

Jessica Davis

Senior Editor

Jessica Davis is a Senior Editor at InformationWeek. She covers enterprise IT leadership, careers, artificial intelligence, data and analytics, and enterprise software. She has spent a career covering the intersection of business and technology. Follow her on twitter: @jessicadavis.

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like

More Insights