IBM’s internal collaboration process is helping its customers in their artificial intelligence journey — from feeding and collecting data to building AI models around that data. The data scientists can gain valuable insights from those models and make critical decisions. 

IBM’s AI Accelerator team does all the internal testing and hands it over the Data Science Elite team for further developing and testing with clients. Automated metadata generation was one of those solutions that started out as a way to save internal manual labor but is now being developed into a real product and presented to chief data officers around the world. 

“We’re doing some automated metadata generation and using AI to — instead of manually having to label and tag that … generate about 85% of our labels internally and drive that into an existing product,” said Caitlin Halferty (pictured left), director of AI Accelerator and client success at IBM. “Our teams really partner well together … in terms of enabling that acceleration and leapfrog.”

Halferty and Carlo Appugliese (pictured right), AI and machine learning program director, cloud and cognitive software, at IBM, spoke with Dave Vellante (@dvellante), host of theCUBE, SiliconANGLE Media’s mobile livestreaming studio, during the IBM Chief Data Officer Summit in San Francisco. They discussed how AI teams collaborate and work together, the new CDO role, client insights, and how IBM protects customer IP (see the full interview with transcript here). (* Disclosure below.)

[Editor’s note: The following answers have been condensed for clarity.]

Vellante: Carlo, talk about your role, and then, let’s get into how you guys work together?

Appugliese: I lead the Data Science Elite Team, which is a group within our product development, working side by side with clients to understand their needs. And then, we work closely with … the global CDO team on best practices, [on] what patterns are seen from an architectural perspective, [and to] make sure that our platform’s really incorporating that stuff. 

Vellante: Caitlin, could you explain [how you work together?]

Halferty: We [the AI Accelerator team] are often the early testers … of some of the capabilities. So, what we’ll do is we’ll test, we’ll iterate, [and] we’ll prove it out as skill internally, using IBM itself. And then, as we build up the capability, [we] work with Carlo and his team to really drive that into a product and … into the market. 

And we share a lot of client relationships, where CDOs come to us, they want advice and counsel on best practices. They’re looking for latest AI applications to deploy in their own environments. 

Vellante: How has this evolved from a theme standpoint? 

Halferty: So data governance and data management, and some of those security access controls, those are always going to be important. But, what we’re finding is [that] CDOs more and more have expanded the scope of responsibilities within their enterprise; they’re looked at us as leaders. 

Vellante: What are you seeing with customers, Carlo? 

Appugliese: We’re seeing patterns around different use cases that are coming up over and over again, and the one thing about data science and AI [is that] it’s difficult to develop a solution because everybody’s data’s different. So what we’re starting to do … is building a platform for these clients, with these accelerators, which are a set of core codes, source code, notebooks, industry models and terms, as well as dashboards, that allow them to quickly build out these use cases. 

Right now, we’re doing client insight for our wealth management and we’re doing that for [the financial services sector]. And they come right out the box of our Cloud Pak for Data platform. 

Vellante: The CDO title has evolved. What are you seeing in terms of adoption of that role and its impact on the organization? 

Halferty: What I’m seeing is, outside of North America, a lot of activity and interest in creating and enabling a CDO-like capability, data leader-like. And some of these guys, I think, are going to leapfrog ahead. And in parallel, those traditional industries. There’s new federal legislation coming down by year-end for most federal agencies to appoint a chief data officer. I think there’s a great opportunity in those traditional industries and also … outside the U.S. and across non-traditional. 

One of the ways our teams really partner well together is we can source some of this acceleration and leapfrog [over the traditional best practices]. 

Vellante: How your customers are reacting to the [digital trade and public policy] framework? I presume “the protect the algorithms and source code IP” is near and dear. They want to make sure you’re not taking models and then giving it to their competitors.  

Appugliese: Every company is a little different on what they’re worried about with that. But [with] many banks, we give them all the IP to make sure that they’re comfortable — and especially in financial services. But, some other spaces, it’s very competitive, and they’re not as worried about it because it’s a known space. A lot of the algorithms we use are all open source, they’re known algorithms, so there’s not a lot of problems there. 

Halferty: And we, as a CDO office, took ownership of that [compliance] across the business and got it where we needed to be. And so, we often encourage our clients to take ownership of something like that and use it as an opportunity to differentiate. 

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

Photo: SiliconANGLE