Beyond the Battlefield — Col. Kris Saling’s Data-driven Path to Transforming Army Recruitment and Well-being

Work-life balance is a complex issue, and the debate between remote and in-person work is polarizing. Treating people as whole individuals with their families, education, and health as variables in a complicated equation is far from being trivial. When applied to the challenges of recruiting staff and soldiers for the U.S. Army, it becomes an even more complex system to navigate.

Col. Kris Saling, the Innovation Director of the U.S. Army Recruiting Command, is well-versed in complexity science as she tackles these tasks using cutting-edge techniques, analytical tools, and data-driven methods. Her career goal is to create the best talent bench for the Army and continually improve the work environment for people. She is passionate about people and dedicated to enhancing their quality of life and work environment. She strongly advocates for remote work and leveraging modern technology to improve productivity, as it allows individuals to participate in nationwide events while being connected with their teams.

In this interview, Col. Saling shares her insights on various topics. These include her career journey, leadership dynamics in the technology space, challenges in defining seniority within emerging data teams in the Army, and the impact of data literacy in building high-performing teams. She also offered a sneak peek into her upcoming book “Data-Driven Talent,” which will be released in August.

Edited excerpts:

The role and the journey

Q: What do you do as a Director of Innovation with the U.S. Army Recruiting Command? How did you get to that point? 

A: By accident! I graduated from West Point in 2001, almost 23 years ago, and I was going to be an engineer. But coming into the personnel space, I started getting into talks about seeing what people in academia and industry were studying in this area.

I was looking at some of the technology I had used in previous assignments for analytics and trying to bring more of that flavor. I have had the support of some amazing leaders.

We started pushing the envelope and asking ourselves some questions about people more holistically — about their ambitions, their potential, their aptitudes, their likes and preferences, the things that incite them to learn particular things, and how they learn.

We pride ourselves in the Army as a leader development institution and are good at it. So I started digging into different ways of recruiting; we set up the Office of People Analytics (OPA), and started the U.S. Army Human Resources Command Innovation Cell. I needed a cross-functional team of creative technologists and subject matter experts who understood the business processes. Then, we brought in operational psychologists and researchers for Peak Performance Management physiology.

We wanted to look at these different dynamics because they all come together with a person in a workplace. In doing that, we started experimenting and the new ideas turned into an Army business process.

About October-November last year, when the Secretary and the Chief of Staff in the Army announced their plan to transform recruiting, that plan included innovation. I ended up here because they told me, "Okay, certainly put your money where your mouth is and build this thing." You can't say no to that.

Data literacy training and beyond

Q: You mentioned learning and training as critical to building successful teams. What training brought the most significant ROI on the invested time and effort?

A: It was our data literacy course; it was designed by a friend of mine who's a professor of Mathematics at West Point. We taught it in person, counter to all of my arguments for being digital-first because it was very much that change management layer. Many who started this course thought, "Oh, data! That's something other people do; I don't use any of this." And we'd have to bring them in and create a comfortable environment where they could ask questions.

We started showing them that they already used data in many ways in their lives, and made them comfortable with some standard definitions and rules. The data literacy course had our highest ROI because it created our data ABCs. Once people have the foundation, they can start plugging into other resources, universities, and self-driven learning programs.

Much of the feedback we got was newfound curiosity about data and unveiling it as something present in everyday operations rather than an obscure concept coupled with a desire to learn more. We are figuring out how to keep capitalizing on that.

I took my cue from Valerie Logan, who runs the Data Lodge, where she doesn't just certify all the individuals who have been trained on all things data but also certifies the trainer in the program. That's what we're looking to implement in the Army.

When building such programs, we strive to answer questions about what the training program looks like, who is running it, at what level of the organization, and how we democratize it as much as possible. The above becomes important because each organization will have different missions and need other types of data and use cases to drive decisions.

Q: Bringing people up to speed with data literacy is no longer a question. It's a must. What about more advanced skills needed as the bar rises?

A: That is something we've been toying with. We should look at it as tiers of customer service. Tier zero is your front-facing interaction and you want that to be as automated as possible. As with chatbots, you want to keep pushing them as much as possible to the non-expert self-serve user. That's where we want to go with our bottom level; we want to keep pushing for technology and advancement to raise the lower level of our non-data professional data citizens.

That way, we free up each subsequent level of data professionals to get more into this field. We've watched this happen in the Army. With more and more technologies, more is needed to understand the math and the design of the underlying system; you have to code and be hands-on. Some of our most technical people at Carnegie Mellon at the AI Integration Center are building highly complicated models. They're pushing the limits of this space.

We want to have people within the branches with the right subject matter expertise do a little more self-service and do analytical work on their own to utilize those advances. So, we're trying to integrate such training into their professional military education to pick up more of that analytical responsibility. Over time, we will shift more of that down to things we consider highly complex now and that a properly trained officer can perform, and it won't be regarded as complex anymore.

Challenges that come with the speed development of new technology

Q: What is the most challenging part about AI moving at its current pace? Not just AI; we have shifts in finance due to the emergence of cryptocurrency, and we are also witnessing leaps in computation thanks to quantum computing.

A: Quantum computing will be a priority as a tremendous game changer. I hate to say it but we're behind the power curve, but everybody is. We've got to get people to start learning the technical side of things as soon as possible. We've got people working in technology and studying science and these processes who still need help to keep up with everything happening.

There is no business field where it isn’t a critical part, where people aren't using generative AI for something, and where people aren't using data and analytics. It's everywhere. I love that we're integrating basic data training and different levels of data training throughout organizations. But we're catching people up to something moving ahead by leaps and bounds daily.

That's the hard part because of how we are going right now; we continue to elevate the entry level to a point where it will eventually become inaccessible.

On using latest tech

Q: How can the latest tech be safely used in everyday operations, especially in highly secure environments? Do you have good practices that you've implemented? 

A: Our secure environments tend to be very safe. In the people space, that's always my worry because our information is inherently unclassified. But just because it's unclassified doesn't mean it's not sensitive.

So, we have to face things from a future where we see the world will be more transparent. We have to assume the information is out there. What are people going to do with it? How are we going to protect it? How will we counter deepfakes, or how can we establish trust with accurate information? What are those trust cues? We're moving into an era of increased transparency where we have a chessboard; you can see all the pieces but you still need to know how they will move. We have been working on these things and building out these types of scenarios.

So, it comes back to people being educated on how to interface with that kind of technology and even with others using technology everywhere. We must understand that the real adverse actors will be subtle, much like social engineering. That's the most significant weakness.

If something is obvious, it will be "I'm not clicking your link. It's come from this gobbledygook email address for something I've never ordered". But when it comes from something that mimics a real person or an actual website we interacted with, that's a different story.

Seniority levels within technical staff vs rank within the Army

Q: You have a wealth of experience working with talent and high technology in the Army, where defining boundaries between roles and transitioning individuals between levels can be challenging. How can you determine if someone has reached a senior level in this context?

A: That's been extremely tricky because the Army likes to default to rank, but we have a lot of authority where we can bypass some of those and promote people based on merit. So we can take them where they are, they have the skills and the leadership potential, and we can put them in positions of increasing responsibility.

The concept of seniority will be function-driven rather than career-defined. Now, somebody has to be the person to say what direction we're going in strategically, so you've got a director with a strategic focus who understands enough of the technology to be able to rally everybody toward that particular vision. You may have some senior technologists on the team who are leading programs.

There's an interesting dynamic where we see not just one leader at the table; nobody's just sitting at the head of the table making all the decisions. You'll have that person with the strategic direction trying to drive the team towards achieving objectives for the organization; you're going to have the senior technologist trying to achieve their objectives for those projects.

I've modeled many teams within the technology space based on the functional dynamics of how special operations teams operate. They're designed with a leader to point everybody in the right direction, but every function in that space is an expert in a particular area, whether it's demolitions, medicine, or cyber communications.

They have to have that fluid dynamic of leadership, not to have one person be the sole person driving everything, to be able to defer some elements of leadership to the person who has that senior technology experience, even if they're not senior in rank, there's a point where rank becomes a whole lot less important. We're witnessing that right now in the technology space.

The outlook into the future and what will be in demand next

Q: What is going to be in demand next? What skills do we need in the next five years?

A: That's always a hard question. We have tried to forecast this many times and have almost a perfect record of being wrong. It will be more of the soft skills necessary in technology — curiosity, imagination, and the drive for continuous learning.

It's like an emergent behavior out of a complex system; it's tough to predict where it's going to come from. We do know that we're driving toward a mixture of technology and human use; that's where the large language model explosion came from; it came from a combination of the technology, the compute power being ready to go, and the human interest in being able to interact easily with the machine.

We see the proliferation of low-code, no-code solutions; we have people who don't necessarily know how to build the code themselves but can drag and drop those modules, becoming more self-servant. The ability to take a lot of this advanced technology and make it relevant for more people will be important, as well as express yourself clearly.

It comes back to the computer's literal nature, where it will do what it's told. Suppose you ask it to do something very different from how you envisioned it, and then the result will be different from what you expected. I think yes, being able to express yourself to other humans and being able to express yourself to a machine, especially as we start seeing more human-machine teaming and collaboration.

About a recent keynote

Q: You are an active member of the data community. You speak at many places and give keynotes. What was one of the latest keynotes about?

A: One of the latest I did was Information Operations Research and Management Science. It's a professional society for operations research analysts, including the Army. I like to put together content relevant to digital transformation, the integration of AI, and emerging technologies.

This time, the organizers asked me to speak more about the practical aspects of those topics. I discussed the human dynamic of digital transformation. I used a storytelling framework we teach in our Data 101 course where we use plot evolvement diagrams, where you've got the exposition, the rising action, the climax, and the resolution. I make my data scientists plan their briefings using that framework. So I plotted that out deliberately as my agenda, built it up, and told why this matters.

I spoke about how looking at the entire transformation spiral, we go from people to platforms, to processes to culture, and how people use platforms for processes within a culture. Those things must be part of your transformation plan for the transformation to succeed.

Another interesting thing that came out of that talk was the idea of change management from a position of empathy. There's always somebody who's not been on that learning journey with us to understand why the new things are important. And if we don't do it from a position of change management and understanding that person's experience, they won't adopt it.

Our whole idea was to tell all these things, this transformation journey, as a cohesive story that brings along all the different players; even if they're not there initially, we pick them up along the way. We catch them up, build them up, and make sure that their experience and their way of using this new, whatever it is we're implementing, whether it's a process or it's a platform.

About the upcoming book

Q: If you had ChatGPT summarize your book “Data-Driven Talent,” what would that summary be?  

A: If you could stick a USB into my brain and I could download everything I learned about talent management, data, and analytic processes in HR over the past eight years and put it in a book, that's basically what it is.

I've been working within the Army people enterprise in various roles for eight years now, and the whole time there, I've been studying new ways, new business practices, and new efficiencies that we can bring and turn into different strategies and projects. It's been hard to summarize all of that. So I started writing it down, and then realized that I wasn't writing enough about this; I could put the details in it, which could be a book.

The book goes from the strategic top-down to the details about creating high-performing, diverse, creative, and autonomous teams. It attempts to address the question of how to do that. It gets into the details of the data on who these people are (not just how many of what type) and how they work by themselves regarding preference, workplace flexibility, and performance management.

It dives into bringing in the team aspects, team dynamics, and how we work together using digital tools. And finally, how do we create the environment around them to ensure these teams succeed?

About the Author:

Katya Mijatovic is Principal Data Scientist at Data Society Group in Washington, DC, leading AI/ML initiatives. With a Master's in Computational Science from George Mason University, she specializes in AI solutions, NLP, cloud computing, and MLOps. A data science generalist in the past, Mijatovic has recently focused on the Energy and Utilities industry, developing custom solutions and optimizing operations through AI applications. Her experience spans multiple programming languages and data science tools.

Multilingual and globally educated, Mijatovic bridges advanced technologies with practical business needs. She finds joy in transferring knowledge between her roles as a parent and a data scientist, often writing about these experiences on Medium. If Katya wasn't a data scientist, she would choose the life of an explorer.