Data won’t spontaneously turn itself into new products and services. Analytics software can help, but it won’t save a data project flawed at the people or process levels. Experts have told us over and over that teams have to get data in their bones for it to work. Has the badly needed cultural shift arrived in the form DataOps? 

DataOps borrows principles and practices like agility and iterative improvement from DevOps. It offers a route for companies to start making gains from data without requiring any huge leaps of faith, according to Julie Lockner (pictured), director of offering management and portfolio operations at IBM.

“Traditionally, what we’ve seen with these large data-management/data-governance programs is that sometimes our customers feel like this is a big pill to swallow,” Lockner said. 

DataOps offers an alternative whereby teams can quickly define a small project and align it to a high-value business initiative. Through this manageable first try, they can learn where they stand with data use in their organization, spot weaknesses, and retune their approach for greater success in the next round, Lockner explained. 

Lockner spoke with Dave Vellante, host of theCUBE, SiliconANGLE Media’s livestreaming studio, during the IBM DataOps in Action event. They discussed the practice of DataOps and how it helped IBM launch a COVID-19 response. (* Disclosure below.)

Sprinting, cataloging collapse time to value

Valuable time is lost in traditional big-data projects for business outcomes. If the purpose is to service customers or to sell to them at the moment they’re ready to buy, then many means for finding, organizing, analyzing, and finally acting on data are simply too slow. A main advantage of DataOps is that it collapses the time periods needed to carry out essential data tasks, according to Lockner.

“It’s about taking things that people are already doing today and figuring the quickest way to do it through automation, through workflows, and just cutting through all of the political barriers that often happen when data cross different organizational boundaries,” she stated. 

While vendor shopping sprees usually aren’t the answer, there is one type of software that is key to speeding up the execution of data projects, according to Lockner, and that is data cataloging. The first step in a DataOps project is knowing what data the company has. Is it trustworthy? How has it been used before? Is it really usable for this business objective? These are questions that data cataloging can answer.

Once the DataOps team has the data it needs, it can use it in a project about six weeks in length. “Take that as an opportunity to figure out where your bottlenecks are in your own organization, where your skill shortages are, and then use the outcome of that six-week sprint to then focus on filling in gaps, kick off the next project, and iterate,” Lockner said. 

A team may even have a full-fledged product at the end of a sprint. Internally, IBM leveraged DataOps to release Watson Assistant for Citizens to provide responses to COVID-19 questions. Lockner cooperated with IBM’s Chief Data Office organization to find ways to automatically organize and classify data to build predictive models.

“That’s a great example of how we partnered with a corporate central organization and took advantage of the automated set of capabilities without having to invest in any additional resources or head count — and be able to release products within a matter of a couple of weeks,” Lockner concluded. 

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of the IBM DataOps in Action event. (* Disclosure: TheCUBE is a paid media partner for the IBM DataOps in Action event. Neither IBM, the sponsor for theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: Julie Lockner