One of the biggest changes is the shift from government-developed to commercial or dual-use technologies. 

The government invented integrated circuits. The government invented the internet. The government invented GPS. I can go on—the government invested in the space program. And I think the government continues to have investments in those areas as well, and that's a good thing for the taxpayers and a good thing for our economy and a good thing for our leadership.

However, today we are much more reliant on commercial technology from a price/performance basis as well as our desire to create more modular systems. With technology today, we see Silicon Valley and others investing in cutting-edge innovations like nuclear fusion, quantum computing, space exploration, and biotechnology that were once the sole domain of organizations like DARPA [Defense Advanced Research Projects Agency] and the national labs. And this has been a good thing as it has accelerated innovation and made us more agile and resilient as a nation.

A big focus for government going forward is filling in the remaining whitespace. Network availability is assumed by banks and Netflix. However, in battle, our enemies strive to take away our command, control, and communication. You must continue to operate in GPS-denied environments and in network-denied environments. That’s why government must continue to lead in these areas.  
 

Two of the biggest near-term opportunities are accelerating AI adoption and eliminating redundant systems. 

In terms of AI, we’ve only begun to grasp its full potential, making this a real opportunity for CIOs [chief information officers] and CTOs to show leadership. There are tremendous opportunities to start across IT operations in code development, system management, and cybersecurity. With our clients, we are seeing real impact here. At the same time, we are seeing increasing maturity for GenAI [generative AI] across the mission space, whether it’s intelligence analysis or claims processing. 

Underpinning all of this is integrating and cleaning your data. Early in my career, I saw first-hand how hard it was to get an accurate picture of operations across the Defense Department. With so many systems, you got any number of answers, and you were never certain which one was right. I’ve been hugely impressed with the success of Advana and its ability to provide that elusive one version of the truth. You can only manage what you measure, and if you don’t have good measurements you can’t manage.

By creating these more integrated data lakes, we can also make strides in eliminating redundant systems and integrating processes more effectively. This is important as retiring legacy systems produces bottom-line savings to invest elsewhere. And the increasing maturity of the cloud and emergence of coding bots are making this more viable than ever. Simply put, the time is now. 
 

They are often relentlessly curious and willing to try new things. They assume that AI can do something rather than expecting that it can’t. And they build in feedback loops. Of course, this type of experimentation in a mission-critical environment requires a solid foundation.  

You need access to the right data. And given that mission requirements are ever-changing, you often need a lot of data. Leaders invest in data collection, integration, and quality. Building competency around synthetic data is a growing focus, allowing them to take on edge cases and manage risks associated with PII [personally identifiable information]. 

AI governance is another foundational skillset, building the right guardrails to secure systems, protect data, and minimize risks. For example, they recognize the need to understand how models might leverage their data and the subsequent need for appropriate controls to safeguard proprietary IP [intellectual property].

Finally, they recognize that we’re early in the AI game and much will change. They are committing to model agility in much the same way that cyber leaders have embraced crypto-agility. This means developing an open architecture that allows them to swap in new models as they evolve and deliver mission advantage. 
 

What excites me the most is moving from static architectures limited to application programming interfaces toward agent-oriented or agentic architectures. This makes your distributed systems even more loosely coupled. It will allow what I’m calling “agent swarming” where you define a problem and the agents within a mesh swarm to solve it collaboratively in the most optimal way. That’s cool.

This is important as it will transform how the ISR [intelligence, surveillance, and reconnaissance] and ERP [enterprise resource planning] systems underpinning government operate. Instead of being locked into a standard approach with static rules, you’ll operate more dynamically, seeking the best outcome for every situation. 

For example, a challenge with monolithic ERP systems is the need to adapt your processes to their business rule. Imagine instead having a mesh of agents that can learn your policies and self-organize to solve end-to-end problems. And if you update your policies, an agentic system can reformulate to readily implement these changes. In an ISR scenario, as new sensors are added to the agentic mesh, they can automatically integrate into a collection process with the right context and in the appropriate order. These will be much smarter and more agile systems. 
 

We are going to see more applications embrace an edge-first approach given the benefits of operating in the field. This is certainly true in the defense sector given the focus on autonomy, but the benefits of smart environments are compelling to other sectors as well. As such, we need to shift our focus from simply operating at the edge to thriving there. Increasingly, this requires seamless integration where applications or agents can adapt to run at the edge or in the cloud, leveraging the advantages of each environment.

AI-based radio access networks—“AI-RAN” for short—can drive the increased convergence of hardware, software, and network needed for this evolution. For example, GPUs [graphics processing units] are used at the edge for wave-forming, beamforming, and load-forming to optimize performance and capacity. In the past, all this was done with “static” best-fit procedures that could not adapt to loads in specific locations based on traffic and network conditions. At the same time, we can run more general low-latency AI given the integration of GPUs at these edge locations, allowing complex inference processing to happen in near-real time.

For federal IT leaders, you need an integrated approach to thrive at the edge. This means driving a similar convergence in your planning with network, product, and AI teams working together on your strategy. 

I'm a huge advocate for open source, and I believe that government should be all over it.

To start, transparency means more secure code. This is the same reason we inspect carry-on luggage. If you can’t see and evaluate the code, how do you know what’s in it? You don’t know—it’s a black box. In contrast, most cryptographic algorithms are open source. This added transparency doesn’t make them riskier—it makes them stronger.

Another reason I'm an open-source advocate is having a lot of people look at the code generally makes the code better. So, you don’t just want open source—you want open source that has active contributing communities.

Finally, open source often provides government with the flexibility to extend products to its unique requirements and get the support it needs to sustain these systems. I think most federal CIOs and CTOs would agree that streamlining licensing and compliance is a good thing. 
 

A key role of any CTO is spotting convergence—connecting the dots across evolving technologies. Take 3D processing as an example. It was once siloed across engineering, art, gaming, simulation, and media, but it’s now converging. NVIDIA’s Omniverse team recognized this early, enabling seamless transitions from physical to virtual and back—like creating a 3D asset and manufacturing it additively. You could argue that this convergence extends to biology, where 3D printed cells are reshaping material sciences. And to simulate such complexity, we’ll need quantum systems that are ideal for modeling molecules. 

Of course, Booz Allen isn’t simply looking to predict the future. We want to create it. This drives our investments across a spectrum of technologies, such as hypersonics, space systems, and autonomy. 

My role as CTO is to anticipate these intersections and set us on a path to arrive at the right time. Technologies like photogrammetry, LiDAR, and Gaussians are merging. Understanding these patterns helps guide both short- and long-term investments. 
 

Maintaining a strong and vibrant partner ecosystem is fundamental to how we deliver technology innovation. For our established partners, we can uniquely provide the mission insight, technical expertise, and strategic resources needed to take on game-changing programs. That’s why AWS named us their federal partner of the year.

We are also focused on identifying and helping to incubate the next generation of innovators through our technology scouting and ventures teams. We believe this focus on disruptive technology is critical to helping the nation maintain its decisive mission advantage. 
 

AI will be the big change. You really need to understand the technology and how to deploy it successfully. Stay close to commercial innovation but also develop your mission- or business-level understanding of your agency’s operations. This will allow you to lead the organization in using AI for real impact. Remember that AI and technology changes are accelerating, so being informed and agile will be the key to success.
 

The technology sector thrives on jargon, so I try to focus on demystifying the technology itself. Put it in the context of overall technology evolution and clearly explain its benefits and risk. Most importantly, let people experience and discover what it can and can’t do first-hand. That’s what we try to do in The Helix, which is our innovation center. 

With AI, we also must recognize legitimate concerns about long-term impact on the workforce. However, if we look at the evolution of software programming—from FORTRAN, PL/1, and Pascal to object-oriented and web services—each generation has generated new opportunities. That’s why I believe AI will be a catalyst for incredible creativity, productivity, and growth.  
 

You must have passion. Hopefully, people feel my passion around technology and the work we’re doing. Passion for the mission. Passion for our customers—that goes a long way. Being customer-obsessed is key.

You must also communicate a lot. I use every opportunity to celebrate our successes but also challenge our current thinking. For example, I do a lot of skip levels, spending time with engineers in charge of specific projects. These creative, ambitious people want to be challenged and inspired by audacious goals.

Finally, I try to lead with humor and humility. I’m not afraid of making myself obsolete and I am willing to dive deeper wherever needed.