HSDF THE PODCAST
The Homeland Security and Defense Forum proudly presents HSDF THE PODCAST, an engaging series of policy discussions with senior government and industry experts on technology and innovation in government. HSDF THE PODCAST looks at how emerging technology - such Artificial Intelligence, cloud computing, 5G, and cybersecurity - is being used to support government missions and secure U.S. national interests.
HSDF THE PODCAST
Leveraging Enterprise AI and the Cloud Ecosystem Part 1
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Welcome to “HSDF THE PODCAST,” a collection of policy discussions on government technology and homeland security brought to you by the Homeland Security and Defense Forum
Our panel walks through how a multi-year cloud migration cut outages, sped recovery, and set the stage for AI that actually helps people at the border and across trade. No hype—just the decisions, guardrails, and workflows that let a 24x7 operation move fast without breaking trust.
Featuring:
- Jay Alalasundaram, Deputy Assistant Commissioner, Software and Applications and Services, CBP
- Baibhav “Bobo” Devkota, Executive Director, Border Enforcement and Management Systems Directorate, CBP
- Thomas Mills, Executive Director, Cargo Systems Program Directorate, CBP
- Neha Saijpaul, Director, Digital Solutions, CACI International Inc. (moderator)
This discussion took place December 12, 2025 at 8th Annual Homeland Security & Defense Forum Border Security Symposium
Follow HSDF THE PODCAST and never miss latest insider talk on government technology, innovation, and security. Visit the HSDF YouTube channel to view hours of insightful policy discussion. For more information about the Homeland Security & Defense Forum (HSDF), visit hsdf.org.
Setting The Stage: Year In Review
SPEAKER_01Good afternoon, everyone. I'm Neha Sejpal. Um, I'm a senior technical director at CACI, and I'm also the delivery execution lead for Regal supporting CP contract. And today I'm honored to be moderating this panel with our exceptional leaders, DACJ, XD Global, and XD Tommills. Um, thank you for being here. So let's dive right in. Um, CDP operates in one of the most complex and dynamic environments supporting mission operations around the clock. So, to kick us off, as we look back on this year, what major accomplishments stand out and how CDP leveraged AI and cloud to support mission needs and the impact it had. Uh, Dap Jay, if you can start.
Cloud Migration Progress And Scale
Biometrics And Mission Use Cases
Trade Systems: Resilience And Revenue
SPEAKER_00All right. Thank you, Neha, and thank you to HSDF for having us over today. Uh, great day of excellent panelists across the board. And for me, it's extremely special to be on stage with two of my XTs, Robo and Tom. And uh thank you for moderating, uh Neha. Uh, yeah, that's a great question to kick it off, right? So AI and uh and cloud. Two of the biggest buzzwords in 2025. Right? Uh it's a two-part question, so I'm gonna tackle it one by one. Let me start with the cloud journey. It's pretty well known, right? I've spoken about it in several events previously as well, about CVP's cloud journey. So it started about five years ago, and uh and our AC touched on it a little bit, and so did Sunil, our CTO, our excellent CTO and CIO. Um, we've been on this cloud journey for a little bit. And I'm happy to say that we've uh migrated almost 90% of our applications over to the cloud, and uh 60% of our database steer. The remaining 40% is not it's not uh chump change. It's about 50 to 60 petabytes of data that's on-premise right now that we are focused on not migrating over to the cloud in the next couple of years. So it's easier said than done, and uh it's a lot of effort and time is gonna go into that, right? And an AI. You asked about AI, right? So the convergence of cloud technologies and AI is gonna be an absolute game changer for CVP. Um and uh I can tell you AI is a resource org, and it goes hand in hand with the cloud technologies and our ability to have our applications and our data decide in the cloud. What we've done in 2025 is uh we've transformed our organization more into like data-driven, proactive systems organizations. So we could we have been focused on how to leverage our data better, and as we are migrating over our migrating our applications over to the cloud. And that's been our biggest focus. And there are several areas that we continue to enhance, leveraging AI and cloud technologies, especially in the mission space, starting with our biometry entry exit, right? It's a partnership that we have with uh our Office of Field Operations and the leadership there. And we've done some excellent work with uh with enhancing some of our TVS, traveler verification service capabilities, and enhanced the process passenger processing as well as our officer experience when it comes to facial biometrics and how we clear travelers, focused on national security and officer safety, but also when when the benefits are the when the officers see the benefits, the traveler automatically get those benefits because they also get processed faster. So that's been a big, big area of focus for us. We've also expanded on uh leveraging cloud and AI in uh touchless fingerprint scanners, leverage uh help helping Border Patrol with encounters in uh in remote areas. Uh we've also uh leveraged Tom and Bobo can speak on it more. In the Ace Trade space, we've leveraged cloud technologies and AI in enhancing some of the capabilities there. Um NII, non-intrusive inspection space, is another one where uh we've done a lot of work where we're leveraging images, leveraging AI to adjudicate images, give the officers the data point that they can use to clear cargo on travelers faster. So uh the sky's the limit. There's a lot of additional capabilities that we've deployed over the this last year. It's been a busy year of uh tremendous success for CVP and OIT, uh, just to name a few, but I'll I'll turn it over to my ex Ds, uh starting with Tom, to touch on some of the ADS IDs worked on from a cargo standpoint.
SPEAKER_03Great, thank you, Dak. Um well, first of all, thank you uh for for the invite. Um it's always good to come out and talk to you to all of you guys, and I see so many familiar faces. It's always good uh to get the IT guy out every now and then. Uh and then I'm not in front of you talking about a system outage or what happened uh and all that. So it's it's very nice to get out and talk to you.
SPEAKER_00That's true.
SPEAKER_03Um so I'll talk a little bit about um some of the things that cloud has enabled for uh the trade applications. Um it really is about resilience, scalability, and most importantly um adaptability to new um data sets uh that help us in our trade enforcement uh mission. Um I I think you guys are familiar. We've done 40 presidential actions uh to date, um, and uh the the revenue collection associated with that is 140% higher than we had last year. Um, and we were able to keep up with all of that uh in large part because of what we've done with our cloud migration bids. Um so we're looking at using uh managed services in cloud. Um, and and I think one of the things that I I'm most excited about that that's coming up is uh starting an effort with the commercial solutions opening pilot where we could actually leverage uh commercial solutions to uh get cutting-edge technologies uh to ingest better data sets to help inform risk assessment. Um, and and what that means is safer goods coming in, more compliance, uh, better revenue, and um and really guaranteeing economic security for for our country. Um so I I I I think that's that's a good summary of of some of the stuff that uh that we've we've been pretty excited about here in supporting our trade trade mission.
Outages Down, Standards Up
SPEAKER_02So Bolo? Yeah, um so over the the last year we've uh you know we've completed most of uh if not all of our journey for our application layer to to the cloud. And and one of the things prior to coming to uh BEMSD, I noticed is you know, there there's these outage calls that happen. Um, and that was kind of the the way of our life for for years and years. And it would be hours and hours of outages. Um it it uh you know used up my nights and weekends, but uh more so it was uh impactful to the mission. And um, you know, the the migration to the cloud has really the resiliency that we're seeing uh from the cloud and uh and the and the speed at which we can react to these uh outages is is tremendous. So having being there now, uh we've kind of set a new standard and and it's a it's a nice standard to have.
SPEAKER_03What a concept we're able to uh enjoy weekends now. I don't miss those days, I'll tell you that.
Balancing Modernization With 24x7 Ops
SPEAKER_01I hope you're not jinxing yourself. You may not know that. You're thinking about it too much, but um that was great. Uh really impressive accomplishments, give especially given CBP's operational tempo. So that naturally raises the question of how do you balance modernization around AI and cloud with the need for uninterrupted 24 by 7 operations across the country.
SPEAKER_00That's a that's a perfect segment to what Bobo was talking about, right? Those uh the resiliency that the cloud offers. It's it's a game changer for us. Uh and I don't miss those days when we used to stay up all night till 2 a.m. with uh with trade yelling at us Tom, sushi ka, sushi brokers and like alpha brokers, right? Sushi, sushi brokers and cheese brokers calling us at 2 a.m. in the morning, yelling at us on why ACE is down. No, we don't miss those days. No, no, it's it's when we started the journey, it started with the governance under our AC's leadership, and uh we came up with the right architecture, and Sunil touched on it a little bit, our CTO, about having the right governance and architecture in place. We didn't want to take that Apple and swallow it on a whole. What we did was we came up with a systematic approach, a roadmap of how do we plan on transitioning our applications to the cloud, right? We drew a line in the sand and said anything that we build new is gonna be cloud-based. That's where it's started. And thanks to our partnership with our Office of Field Operations personnel, our Border Patrol personnel, AMO and our mission partners were supportive of that vision from leadership from top down, right? And that was that's where it starts. And then the next step was to look at our mission to application space, the legacy application space, I'd like to call it, and say which ones could be refactored, containerized and into microservices and stuff, and move to the cloud, and which one had to be blown up and rebuilt. So we had to come up with a roadmap across each of our organizations. And having the governance in place helps us build incrementally and not and have a faced approach to it. And uh, instead of just trying to swallow the Apple on its own and then cost outages all over the place because the architecture is not gonna match the the our efforts for resiliency. I'll tell you, CBP operates in uh 253 ports of entries and over 150 plus border patrol stations. No one port of entry or a border patrol station is alike. So any any impact to those mission partners is uh is a detriment to our national security and our officer safety and agent safety. And we don't want to mess around with that. Or and and the economic viability of a country, like Tom Tom Tamstad, and I mentioned the sushi brokers and cheese brokers. I mean, it's it's a very complex mission. We're truly a 24 by 7, 365-day mission. We don't want to compromise that. So what we did was we we leverage agile principles, agile approach, faced approach. And the most important thing that uh we established under our AC's leadership and our CTO's leadership is uh monitoring and proactive monitoring and addressing vulnerabilities and issues before they occur is the reason why we can actually sleep at night. Right? And that was a game changer for us. And that's that's one of the reasons why we've been successful in this cloud journey, because we wanted to have balance um getting there with making sure that our operations and maintenance is not impacted. Because our mission applications can't take a break.
SPEAKER_01Yeah, yeah.
SPEAKER_00You want to add anything to it? I'm sorry.
Governance, Architecture, And Monitoring
SPEAKER_01I was I was just gonna say, we always used to be DevSecOps, but when when CVP started using it, it was Sec DevOps because we are moving everything to the left, catching things before they become inertia. So once we balance modernization with continuous operations, um the next big thing we are seeing is the speed. XD Bobo, I would like to get your thoughts on this. Um we are seeing AI use cases generate mission value more quickly, sometimes in weeks instead of months. How is this shaping the way CEP thinks about uh delivering solutions? And how are we ensuring that answers are correct, secure, and trusted?
Speed With GenAI And Human Oversight
SPEAKER_02So so you know, kind of at a to answer the last question first, um, we we always have a human in the loop uh with any any AI product or any any tool that we we build. Um but but to the point of of how do we how fast are we able to deliver them, right? So thanks to uh CTO Madagari and and Dak Mays and having the infrastructure and and the the AI ecosystem available, um you know we we uh CTO mentioned earlier about chat cb and it's deployed across the agency. Um as we as folks onboard uh generative AI into their workflow, um, they start coming to us with with targeted use cases. And what we're seeing is because we have the architecture in place, because we have the infrastructure in place, it's it's really it's it's taken that final step. Um and that final step in the generative AI space seems to be things that we could deliver very quickly because of of having those, having the setup, right? So we're really just uh just finishing off that end product for the users. And uh these things are getting delivered very, very rapidly. Um and to your point, we're we're able to turn things around a week. Sometimes it's been I've I've seen them happen even faster than that, uh and so it and the thing is it's it's all of that architecture and having that in place is facilitates that. And and I don't see that decelerating anytime soon. Um it's it's really just a question of when when the mission identifies more and more items that can be um that AI can can be applied to it. So it's only going to accelerate um and it's great.
SPEAKER_01Are you giving me a heads up that you have your next use case ready for me?
SPEAKER_02I am.
SPEAKER_01Um and of course, all the speed data and the new delivery models ultimately depend on our people. So, DACJ, how are we preparing the workforce to effectively use and trust these tools?
Workforce Upskilling And AI Code Assist
SPEAKER_00Yeah, getting those capabilities out means the users want those capabilities, right? So it has to go hand in hand with user training, not just our end users, but including our development teams to be able to get those capabilities out faster. Right? And like Bo XT mentioned, we talk about getting capabilities out within a week or near the next day and some of those capabilities. That comes from resource upskill. Right. We need to make sure our employees, our like IT staff, are one equipped with the right tools to be able to leverage AI and cloud technologies the right way. And uh in the last six months or so, we partnered with industry and our CT organization to equip some of our developers as a pilot with uh code generation of tools to be able to play around with it and see how AI can save some of our time. Our AC will probably have the dashboard handy, which I don't, unfortunately. He's my he's the dashboard guy for us, right? He's awesome. And uh and the the dashboard shows exactly how our developers are utilizing the AI tool, the code gen tool. And I can tell you, just within the last couple of months, we've seen a significant amount of savings. We've seen immediate 40% savings and the amount of time a developer takes to generate code using it and accepted lines of code, resulting in over a million dollars in savings. Right? And that is only a small subset of our development team right now. Translate that to, I know CTO talked about chat CBP and having chat CPP now available to 65,000 CPP users. All of our employees and contractors have access to it too. What we have to do now is how do we now educate the workforce to take advantage of those tools that we're able to give them, right? And that comes with training, communications, outreach, and that's easier said than done. And I think that's an area of uh of focus for us as we transform our workforce, not just our IT workforce, but also working with our mission partners and making sure that they're also able to take advantage of these tools to make their lives easier and take some of the administrative burden away from them. I think that'll be a game changer for CVP and as all because our mission doesn't stop for anybody. Right. I think uh, Tom, you have some specific use cases that you're working on from a trade space that you can address.
AIOps, Legacy Code, And Faster Fixes
New Data Sources And Trade Risk
SPEAKER_03Yeah, and I I like like um Dak J had said, um, you know, the training um afforded by you guys heard from Sunil earlier um and and sponsored by uh by our AC has given us a lot um in the in the realm of AI. And I'll I'll just give give a couple um instances of of where we've seen um a lot of uh um traction in what we could do with AI, particularly in the AI code assist area. Um Niha, you'd mentioned uh about modernization and how do we balance modernization around um uh moving uh to AI and cloud and balancing that against our 24-7 operations. Um, you know, what cloud did offer us is a lot more automation for deploying code into production. Uh, and then what we've also subsequently found out is that cloud native pipelines integrate well with AI ops tools. Um, so it really gives us much better insight into the health of our applications, all the underlying supporting infrastructure. Uh, we don't have to now get on calls with the infrastructure team. Actually, the development teams themselves can figure out um what's going on because we have insight into operations that we never had before. Um, and that results again in in a lot fewer incidents. Um, I I I do know that this year uh every every time we weekends are back now and you get suspicious, like why is the phone not going on? I I know transaction volumes are higher, but the phone is not ringing. Um uh but that that's what we're seeing, fewer incidents. And then when there is an incident, it's much faster uh to resolve because uh the these these ops tools and these AI ops tools give us insight into uh where the problem is rapidly. Um AI has also come to help us in approaches to modernize applications. And we we talked about um AI code assists, and I can tell you um I I know uh you all are very familiar with the problem of uh of legacy code, or what we I don't know if we're allowed to call it legacy code anymore. Um but uh what we found is that using these um code assist agents, we can actually interpret uh code that was written many, many decades ago and help us accelerate um those those cloud migrations. Um it's been a it's been a big deal for us. Um I know uh some of the stats that were thrown out for for the senior developers, uh, they're seeing um a significant increase in productivity uh because the uh AI code assist is helping uh break down a lot of the old code libraries and what was done and how it's written into uh persistence layer. Um, and and it's something that they don't have to go through the millions of lines of code to interpret that. Um and then lastly, I'll tell you um uh what we've done uh in the AI um uh realm, uh especially with some of our newer technologies that we're adopting. We're allowed to, we're we're now able to ingest data from uh what we call non-traditional actors in the in the trade space. So, you know, in the trade space, the traditional actors have always been your importers, your exporters, your carriers, your brokers. Um, with the commercial solutions opening pilot that I'd mentioned earlier, uh, we're able to leverage uh data that comes from commercial platforms that have much richer insight into the uh logistics supply chain. Um, and and you know what that means for for us and CBP, better enforcement, uh better revenue collection, uh better uh regulatory compliance, um, and then for us and the public, uh, you know, economic uh security and and a lot more reassurance that the products that you guys are are buying are are safe.