Artificial intelligence achieved prominence in 2017 as companies looked to automate how they derive value from data. Following years focused on collecting data, where Hadoop and big data management dominated the conversation, organizations are now turning their attention to machine learning and other forms of AI to better extract meaning for that data and to open up new business models, products, and services.

For perspective, 451 Research expects the total data market to reach nearly $140 billion in 2021. Couple that with a rapid rise in job creation for AI skills and you have the telltale signs of big opportunities. Recently, the AI Index report from Stanford noted that the share of jobs requiring AI skills in the U.S. has grown four and a half times since 2013.

To put data to use with a focus on AI, organizations should recognize that organizational change is as important as technological change. With that in mind, I recommend these principles.

Keep it simple

Where possible, try to consolidate infrastructure and systems and automate with DevOps methodologies. It will make more sophisticated workflows easier. Understandably, people will make their own choices; just try to keep them in check through a consistent philosophy and set of methodologies. Work to close operational gaps by focusing on delivering applications to corporate or end users, not just analytics. Analytics remains critical, but applications help drive the business forward. You should also aim to streamline workflows by bringing existing and new data into select systems that can process multiple functions at the same time.

Keep it real time

We live in the here and now, and the pressures for rapid response and access to information are only growing. Bringing real-time data to any organization energizes the appetite for transformation and increases capacity for positive change. For data architectures, this can mean scoring new data upon arrival based on existing machine learning models, such as those from SAS. This essentially takes long batch processes and converts them to continuous workflows, reducing or eliminating lengthy data transfer operations.

Real-time data offer more personalized experiences, more interactive commerce, and the most accurate representation of what might happen next. Companies that can create this line of communication with users stand to benefit from more intimate customer relationships.

Keep it adaptable

The pace of change is accelerating, so naturally modern workloads continuously change. Companies continue to launch new projects and products, and success is never perfectly forecasted. To stay adaptable, new applications can make use of both transactional and analytical functions together. This reduces development and deployment time and shrinks overall infrastructure needs. More importantly, it puts application developers in a position to offer the richest experiences possible for their customers.

Applications that develop a feedback loop with built-in analytics can focus on learned behavior. This can quickly evolve to systems recommending decisions on your behalf.

Examples of putting data and AI to work

Let’s take a look at real-world examples across finance, media, and energy industries.

Finance

Nothing moves as quickly as money and investing, and finance remains a crucial industry for new data application developments. In describing how “Morgan Stanley Is Creating the Financial Advisor of the Future,” brandchannel noted,”One key area in which Morgan Stanley is investing that capital is tech innovation, including AI- and data-driven digital and mobile systems, automated investing, and customized multichannel communications, all with the goal of improving the customer and employee experience.”

Media

Analytics in media is a white-hot space as companies transition from traditional broadcasting to live linear streaming, a major industry initiative. With more consumption through mobile and web applications, media companies need to capture and analyze data in real time and use AI to help make recommendations and decisions.

Viacom noted in its last stockholder letter that it was committed to “Harnessing our Powerful Data Products.” Company representatives explained, “Viacom’s advanced capabilities in data science and research continue to unlock real value for the company. … To further our competitive lead in this field, Viacom also launched Vantage Intent, a cutting-edge analysis and modeling tool that forecasts consumer purchase intent. We believe these products are revolutionary in their ability to provide advertisers actionable insights, and both have the potential to generate future growth for Viacom.”

Energy

In the energy sector, Pacific Gas and Electric is working on a proposal for Grid Operations Situational Intelligence (GOSI) intended to “demonstrate new technologies and strategies that support integrated ‘customer-to-market-to-grid’ operations of the future.”

The executive summary says, “To operate the next-generation grid safely, affordably, and reliably, electric utilities will need to integrate significantly more data and information into both existing and future operational systems.

“And one key accomplishment showcases the need to use data intelligently and in real time, providing a real-time emergency operations center (EOC) dashboard and reporting tool to replace existing Excel-based dashboards. These provide a single-source of truth for data and consistent and timely reporting.”

Getting started with data and AI in 2018

Large enterprises already have numerous applications and data repositories in place today. However, with a broad business focus on digital transformation, the time is right to pursue new data applications and incorporate machine learning and artificial intelligence.

Given the resources available in the cloud from Amazon, Google, Microsoft, and others, it often makes sense to start there first. But companies should keep in mind that the ability to deploy solutions on any cloud or in their own data centers may be an important prerequisite to optimally service their companies long term.

Remember to start simple, perhaps with a real-time dashboard. From there you can move to an externally facing application that incorporates a machine learning scoring model. At that point, you are on your way to more sophisticated AI applications.

Gary Orenstein is the senior vice president of product at MemSQL, a real-time data warehouse for cloud and on-premises that delivers immediate insights across live and historical data.