The Department of Health and Human Services Chief Information Officer Jose Arrieta took over as the department’s permanent top IT official in late May 2019. Arrieta has become a champion of embracing emerging technologies and is currently working on several new ways to harness technological advancement to benefit both internal and external customers. Federal Times reporter Andrew Eversden sat down with Arrieta to discuss his successes so far, his priorities moving forward and why the health care space is ripe for technological risk-taking.

You’re four months in as HHS CIO. What progress do you think the agency has made with you?

WELL, I THINK THAT one of the interesting things about any job, whether it’s CIO or senior procurement executive, is there’s a lot of good work being performed throughout these large organizations. And the interesting thing is the progress never stops. So, we have made a lot of progress in ... beginning to consolidate our governance processes. We have made significant process in starting to establish and brainstorm pilots with other agencies.

DISA [the Defense Information Systems Agency] is one of the agencies that comes to mind. From a cyber front, we’ve built a neural network [algorithms designed to identify patterns] that is focused on structured data, and from a non-cyber front, when you look at our statements of work and at the prices paid in terms and conditions, we have actually created the capability to use a recurrent neural network to read unstructured text to vectorize that text.

What’s an example of that?

“THE CAT IN THE hat, the cat wears a red hat” — it’s actually taking those sentences or reading them, turning each word into a number, and creating a vector and pushing it into a predictive model and starting to look at labor categories and [position descriptions] to get an understanding of what labor categories and job descriptions actually align. And then, using that as a stream to actually compare price points on labor categories, as well as the price point associated with hiring a federal employee. And we’ve been able to do that very rapidly.

I’ve focused in on three or four pilots:

One is HHS Accelerate [a new acquisition portal that will use blockchain and artificial intelligence to save the government money on contracting, similar to price comparison tools used on the public internet].

Then there’s the recurrent neural network focused on reading unstructured text data, vectorizing it and pushing it into a predictive model to get an understanding of prices paid on professional services labor categories.

There’s also a neural network focused on structured data to understand the threat landscape associated with what hits our trusted internet connection or TIC environment.

Then, lastly, brainstorming this idea with DISA of using a behavioral-based identity to establish identity for an individual that is accessing our network, so using 240 behavioral characteristics versus using a username and password, which we think will be transformational for our first responders in particular.

I am proud of that progress.

How is it affecting your workforce?

I THINK THAT THE other piece that I would say is that, while this organization has always been customer-focused, we’ve taken it and done a relook. We have internal customers that we service and we’ve reengaged with our customer base to develop an understanding of what their needs are and how their needs may have evolved. When you work at a federal agency for a long time, especially a large federal agency like HHS, what you realize is that there’s a lot of turnover.

Even if you’re doing a great job with customer service, you know a percentage of the people that you’re serving are always kind of moving on to different jobs and you constantly must engage in order to maintain that level of customer service. So that’s something that’s never stopped and we’ve kind of refocused in on that to hopefully drive better value for our customers.

What capabilities will HHS Accelerate give you in one year that you don’t have right now?

SO, WHEN YOU THINK about Accelerate, we’re thinking about serving different personas.

One, if you’re an industry partner, our goal is to lessen the burden on an industry partner. Imagine you’re an industry partner bidding on a federal contract. You have to submit [representations and certifications], have to go through the validation verification process. You have to do a financial. Many agencies will do a financial responsibility determination on the work that you’re performing. You have to resubmit SAM.gov information. If you’re a large company, every time you submit an RFI response — for analytics, let’s say — at HHS you just have to redo all that work of creating that RFI and submitting it.

What we want to do with Accelerate is automate those tasks. We want to lessen the administrative burden on industry so they can focus on the critical value that they add, which is the thinking, the capabilities that they bring to the table. So we want to automate the SAM.gov pull. We want to automate financial responsibility determinations [so] all you have to do is update your numbers. We want to automate reps and search checks. We want to automate validation and verification. We want to create an immutable record of your last RFI submission to the agency. So when anybody from your company logs in to submit another analytics response, an RFI response to an agency like HHS, they have a record of what was submitted before and they can just build off of that, tweak it or change it. So that’s kind of one person we want to service. That’ll lessen the administrative burden on industry partners, allow them to focus in on helping us solve our most critical problems. We think that adds a ton of value.

The second persona I’ll talk about is program managers. From a program management perspective, we want to empower them. We want to enable them with the information that they need to do their job. So the idea of the recurrent neural network that, as I talked about earlier, we actually have built it to merge Statement of Work information. So let’s imagine that you’re a cancer researcher from the National Institutes of Health and you come to HHS to do cancer research and you’ve been given $10 million because you have an idea that you think can solve a problem. You know that before you can even start to expend that work you’ve to go build an acquisition plan.

Now we have some of the brightest smartest cancer researchers on the face of the Earth here at HHS — and multiple other disciplines, as well — but they don’t know how to write an acquisition plan. Is there an opportunity cost if I’m focusing their energy on [acquisition regulations] versus focusing their energy on: what is your problem?

What if I could, just like TurboTax, ask a series of questions and then take the historical information associated with cancer research across the department and pull it to the forefront so that cancer researcher had kind of like a market research report that gave them information as to what occurred in the department before. And it also included information from the vendor portal, from information that industry submitted so they had a look at kind of what industry partners could possibly perform their research. And then they could read that — they were enabled to actually read that market research report.

The last persona we focused on is creating value for the contracting professional. You know when you walk into Target right now, or you walk into Best Buy, and if you actually look up a price on Amazon and you show the price on Amazon to the cashier or any other website and it’s cheaper than the Target price or the Best Buy price for the same product, you get an immediate discount.

Why is that important?

THAT IS EMPOWERING. So we want to empower contracting professionals with that insight and information. Does it mean we’re going to make a decision for them on prices paid in terms and conditions? Does it mean we’re going to make a decision for them? No. But what we’re going to do is we’re going to give them information so they can focus on the critical thinking skills required to actually negotiate the prices paid in terms and conditions. So we think that that will improve the amount of time that they have to focus on critical thinking.

You’ve gained quite the reputation as someone who pushes emerging technologies and digital transformations. Why is health care the proper place for federal dollars to be spent on that kind of thing?

WHEN YOU LOOK AT global expenditure on health care, the United States is 55 percent of the global marketplace. That is huge. We have a significant population of people that have a need for health care.

Health care is the only market in the world that touches you from the day you’re born until the day you die. So, from my perspective, where else could we make an investment that would have a bigger impact on the lives of American people?

I think that’s why it is extremely important, whether it is emerging technology or just standard blocking and tackling, to actually improve the way we deliver services for health care services for the American people.

Over the next 18 months or more, what are some of your priorities and other places you’re seeing opportunity to invest in these emerging technologies?

WE WANT TO TEST [HHS] Accelerate’s integration with the financial system, then fully deploy it. That is one of our goals.

We want to actually test behavioral-based identity at the edge of the network. We’re brainstorming at this moment to figure out how to do that. But I want to be able to use 240 behavioral characteristics about you as an individual in terms of how you use your wireless device or data that’s taken off a wearable and I want to be able to use that to give you access to the network.

And I think that’s very empowering, because no longer do I need to use username and password to give you access to a network. I can use how you behave. I can use a retina scan, I can use a facial scan, I can use a thumbprint, I can use the fact that you connect to your secure Wi-Fi every morning at the same time when you wake up and that you’re also connected to your Whirlpool dryer for whatever reason.

I can have a behavioral record of how you use your applications and who do you text and how often do you FaceTime with your mom and how often do you just phone call with your girlfriend. What is your gait? What is your heart rate? What is the humidity outside? What is the temperature?

I can use that to establish a series of behaviors and a confidence score that says, “Yes this is you trying to access your phone.” I think that that is extremely empowering.

Imagine that you’re working in an operating room in the health care sector and the way that you access your systems currently is you leave yourself logged in, because if you forget your password there’s no time to reset to log in to the system. If you’re using behavioral-based identity as a mechanism to create access, you can have immediate access based on a series of behaviors that are occurring.

And then, when you don’t need access you can lock the system. We’ve improved the ability to control cyber risk as it relates to the way you work. I think that’s extremely powerful. I also think it’s extremely powerful for first responders within HHS. We have first responders that receive multiple data streams. Imagine, instead of logging into multiple devices and having multiple usernames and passwords, using behavioral-based identity at the edge of a network to give them access to the data sets they need to do their job.

It’s throwing a small rock into the ocean ... And we’re doing it, we’re testing it at very low cost, we think the ripple effect will be huge.