Episode #37: Tracking Trends While Looking Ahead: Federal Software Factories 

  • we@designindc.com
  • December 21, 2023


Announcement: Broadcasting from Fairfax, Virginia, you are now listening to The Highlight Cast.

Ashley Nichols: Hello, and welcome back to Highlight Cast. I’m Ashley Nichols, Highlight VP for Corporate Strategy and Development, and I’m happy to be back hosting as we discuss the future of software development. We have Rise8’s Prodacity event. Uh, earlier this month, uh, and there was some amazing insights and conversations around development and procurement, um, how to push the envelope and development, the federal sector, and it inspired today’s conversation around current trends, development, software factories, and innovation, so welcome, and I am pleased. To be joined by Highlight VP of National Security Solutions, Kevin Long. Hello, Kevin. Hey, Ashley. It’s good to be here. I won’t beat around the bush. We’ll, we’ll get right to it. Let’s do it. What are some of the major takeaways from Prodacity ? Uh, that you’re most excited to apply to our work here at Highlight.

Kevin Long: I was actually really surprised with, uh, Prodacity with the way they, uh, put it together this year in the best possible way. Um, where they had it in all around geeky stuff, but they had, uh, culture, technology, procurement, sort of different tracks around, around that. And I went to some of each, um, but when I took away most of it was, uh, Two things data and culture and how they interact. And then also, I mean, ubiquitous AI everywhere. They talk about all that, but, you know, culture in terms of how you implement AI and what you want to do with it, how you can do it, do it intelligently, the data that drives good AI and things like that. But really, a lot of it. It was, you know, how, how, you know, you have the right, the right data, the right information, and how you have the right culture to encourage innovation and change, uh, where it’s necessary. And so those were the big things that I took away. 

Ashley Nichols: Yeah, I was, uh, I was busy at the booth for a lot of it, so I missed some of the keynotes that were, I felt like a lot around leadership and culture, um, in the environment. I really enjoyed the heavy participation of government personnel. And really getting their insights. I think a lot of these conferences, you don’t, you don’t always get that right. Yeah. A lot of times it’s echo 

Kevin Long: chamber from contractors for sure. 

Ashley Nichols: It is right. So really hear some of their needs and challenges. And what they’re trying to do. And I think what I found was a lot of it’s fairly universal, right? People are coming up against the same challenges and roadblocks and, and looking for similar solutions across those sets, which, you know, for someone like us is super valuable. Right. But I also really enjoyed that. They had a whole procurement track this time. Uh, yeah. And so I spent a lot of my session time there. So, yeah. Yeah, it was really well put on. It really was. Yeah. So, with that, what were some of the top challenges that the, that you heard the federal government were facing with regards to soft 

Kevin Long: result? I think about, uh, some of the top challenges that they’re facing, uh, around a lot of what it is. people chose to highlight as their successes, right? Because if it’s easy stuff, no one wants to talk about it and hear about it, right? Because it’s always been done. Uh, and you don’t have the wow factor. Um, and so the, the things that really stuck with me were data sharing, uh, stuff that VA was able to do that allowed them to have APIs being shared out to, uh, remove a whole bunch of siloed information. And they were able to implement that through, because it’s a RISE 8, uh, thing, continuous authority to operate, right, with CATO. They were the first, uh, FedSiv agency that, that got their, their CATO to be able to do that. And so, uh, really, you know, keeping your data secure, making it, making it visible where it needs to be, but only to where it needs to be, uh, and being able to see all of the data and information that agencies have, because they’re They’re collecting it at a phenomenal rate, but they’re, but if it’s not shared, if it’s not secured, and if it’s not accessible, then what does it matter? And so that’s really the, the, the challenge that I think that they were really focusing on. 

Ashley Nichols: I think that matches up with a lot of the procurements that we’re seeing coming up to write a lot of the DevSecOps, which you think about as. More systems, case management, you know, that kind of stuff. A lot of the DevSecOps procurements that are out there now are around data, around AI, right? And I think it’s really governance, right? And governance is the thing that really focuses on, um, it really tells you where we’re at, right? Uh, with, with the, with the upcoming needs within the federal government. This was largely, um, DoD. I would say there was a lot more DoD folks here than some other folks. Uh, so it was a bit more software factory focused. 

Kevin Long: So, so DOD and law enforcement, it’s like, if, if people that you serve probably need to carry a gun at some point, they were there. So we saw some law enforcement folks too, but yeah, yeah, 

Ashley Nichols: yeah. You know, it’s the criticality of information in the field really is what that really all comes down to. And that speaks with data, uh, and the AI, um, and the mission criticality of the systems that we’re talking about. Um, Some of the trends that I think that we saw focused on, uh, in the discussion were, I’ll start with procurement flexibility, and this is, you know, where I really kind of dug in. It was a big learning experience for me, you know, familiar with things like OTAs, right, other transactional authorities, um, and SIVRs, you know, uh, innovation, R& D. Request, but they’ve talked about, uh, broad broad agency announcements as well as CSOs. In terms of these different ways that they’re procuring what they mean, and a lot of times it’s not even on a large scale, you know, it’s the desire to prototype innovative solutions into their spaces before they go sort of whole hog in a direction, which I think is, you know, Really critical to the speed of fielding these systems. I think I actually saw, um, I can’t remember where it was, but it was, uh, you know, someone, I think I have an air force that was talking about when you’re preparing for the next challenges that we have from the military standpoint and He was referring specifically to China, you know, he says that we will be ready by 2036 when really we need to be ready in the next three to four years, right. For the kind of cyber challenges, right. That, that, that, that, uh, relationship brings up. Um, and so I think that these alternate procurement strategies are really focused on, you know, Being able to speed up that timeline, because we all know that the process procurement process right now is, it’s always been slow with the idea of getting the best value for the government, which is an excellent aim. Um, but in a, in a, in a time of protest, it is just really kind of ground to a halt in terms of being able to, you know, get technology to the warfighter. Um, which is the real focus 

Kevin Long: Yeah. Getting the right tools to the government in a timely fashion. Uh, yeah. It was, um, uh, the CTO at a, at the Bestin software factory, who was a procurement officer at Kessel Run, did a really great. Presentation on avoiding, uh, competition theater also, um, around procurement and, uh, the number of, uh, contracting officers that were in that room listening to, to ways to make sure that they’re asking the right questions, that they have the right, uh, the, the right requirements down, that they’re, they’re procuring it in the most effective way. I mean, because like contracting officers, I mean, they hear co challenges or they hear orals. Yeah. I mean, you know, they know the far and the D far, right? And so, uh, they’re listening to other folks to try to put it in the right place. And his, his talk around, uh, around that I, I thought was really informative from a, uh, uh, hearing it from, uh, Former Fed contractor point or Fed contracting officer.

Ashley Nichols: There were a number of former federal acquisition officials that were there that are now focused in the private sector, but very specifically on focusing their former. You know, on helping their former colleagues get what they want in terms of procurements. Um, a lot of, you know, consulting facilitation for both the government and industry too, because what was also interesting about Prodacity was that a lot of the industry participants were small businesses. I did not see many of my normal large, large business cohort of folks. I usually see, um, we’re large business. But, uh, we’re definitely on the emerging side of that. Uh, so it was, it was an interesting talk at the end. Dr. talks to a lot of folks who are skilled similarly to us and in sort of the same boat as we are and talk about how they are trying to service their customers, the agility they’re trying to achieve. Um, and I think that speaks a little bit to, uh, where the, the target. The target audience, but where a lot of the people providing these Agile solutions, these truly Agile solutions are on the smaller side, because they do offer that kind of procurement. It doesn’t take away maybe from the scale that you need for some. You know, a lot of large system support, especially established system support, but when it comes to, um, 

Kevin Long: But solving the new problems, you can’t, you can’t put it into a, you know, 75, 000 person machine and expect it to move quickly, right? 

Ashley Nichols: Yeah, and, and some of those, uh, alternate procurement, Vehicles are open only to smalls, you know, right? Um, you know, trying to, you know, get a twofer. I also read that the small business engagement across BOD has declined considerably over the last 10 years. Um, Yeah, yeah, absolutely. This is a problem across federal government that they’ve been trying to reach, but I think that, uh, leaning into innovation is, is a way specifically that the development community, uh, is really trying to bring some of those, those skills back in, um, To the collective government. And interestingly, I think the club club, we have a software license management solution called Atlas that we, that we, you know, had there with us that got a lot of interest, but I will say that it’s because, you know, while we develop a firm, one customer specifically, um, it’s software license management, software asset management is continuing challenge throughout the development. I think that there are millions of dollars of assets and licenses that different agencies can’t account for. 

Kevin Long: Underused or paying overages because they can’t predict what’s going to happen. Uh, what’s going to happen? Don’t have the ability to predictively model. Uh, what if scenarios around if things scale up or scale down aren’t able to negotiate better rates because they don’t know how many they’re going to buy of any particular license at a time. Yeah, it’s, it is a, It’s sort of like a meta problem for like software factories and large software software groups. It’s when I talk about, I mean, I have a talk on on the foundations on building software factories, but the foundation of it, it’s having a tool like Atlas that again, it’s, It’s about the data, right? Uh, so people, so people know what they have, who’s using it, when it was last used, do they need to have more? When did the terms and all of these, all of this information so that, I mean, a software factory or a branch in the military or, or any other federal government is like a business. They have a budget. They have to run it. And if they don’t know what they’re spending and nobody wants to pay attention to Gmail licenses. Jira licenses, right? That’s yeah, it 

Ashley Nichols: is not the sexy. 

Kevin Long: No, it’s 

Ashley Nichols: work, right? But it is a critical. 

Kevin Long: It is the foundation upon which all the other it happened. You can’t deliver the software if you don’t have the tools and having sort of just Top notch, easy to use, uh, data forward, uh, visibility into the tools and the assets that you have at your, at your disposal is, is critical. And it is, uh, it’s, yeah, we saw a lot of great interest In hearing about it, because while it’s not the, Hey, we’re reprogramming how to refuel, uh, uh, uh, bombers, it is those guys can’t program how to refuel bombers without the tool that we provide. Right? And so, yeah, so I’ll claim credit for that. 

Ashley Nichols: Also, as we see some consolidation in software factories, as well as, software factories are established as a concept throughout the DoD, right, and, uh, depending on who you talk to, varying degrees of success depending on, um, a lot of things. So some encouragement is like, do we have too many, do we need to pull them back in? And as you start to consolidate these things, do you have sort of a multi tenancy situation, right? Or multiple customer situation. A hundred percent. 

Kevin Long: You just got to where I was about to go. I love it. 

Ashley Nichols: Yeah . Which is where you really also the visibility into those things. Um, from a financial, from a FinOps standpoint, uh, you know, becomes critical and the capability to do that. So that is, uh, you know, certainly I need the FinOps software license management is inextricably linked to the FinOps of these organizations. Right. So. Um, we’re seeing that need pop up. And again, you see procurements around that, or they’re embedded in a lot of those needs are embedded in a lot of the procurements, like Air Force One and the different CFP. Um, 

Kevin Long: yeah, well, I mean, any, any, any organization that has multiple teams with different missions, Or is, uh, like you said, it’s like literally multi tenancy where it’s fee for service or things like that, right? Or any, any horizontal, uh, that’s doing that, that has to support multiple different missions. There’s the Venn diagrams of, of IT overlap are. Get really complicated in terms of who can use what for, for what tool. And, you know, what tools do we already have good licensing deals with that do the same thing as this other thing that people are looking for. And so knowing what, what you have in, in your quiver and, and can roll out and do it affordably and, and manage it all together is. It is, I, I don’t want to say this is, it’s like a lot of these groups become like innovation silos because they sort of had to cordon themselves off a little bit to be able to operate on, on, yeah, on a finite set of this is what we’re going to work on and then we’re going to get it out. Um, and so with that, the more you do that, then. You know, you have a hundred different things all doing that, but then you’re siloed again. And, you know, so as you were saying is, you know, there’s some consolidation or even, you know, up a level at like a CIO that they can then provide down to all of that across all of the multi tenants is, you know, Absolutely. The next step, both in cost savings, understanding, and, uh, the ability to provide the right tools for, for the software. 

Ashley Nichols: Yeah, that’s a good point. It certainly applies to any CEO organization who’s responsible for supporting the multiple tenants of their agency. Right. Absolutely. Um, and you know, all CEO organizations are not necessarily created equal, some are more guidance and some are more, we actually control the tax, uh, but for those. Folks for each office, you know, they they constitute a different tenant with a different budget um And they need the same accessibility uh consistency of stacks and visibility for Finances. Absolutely. Um, one other interesting thing is and I won’t go into it here because I don’t think anybody’s really felt it yet Was a lot of it cross agency interest in being able to procure from other services meaning Coast Guard, buy from Army. Cloud. Air Force, buy from Army. You know, uh, in different areas. And, and I heard tell of somebody who, Made one of those things happen. But that is, that is a real challenge that takes, I think the political will of both leaders and those two services to be able to come together and make that happen. But, uh, that kind of, uh, certainly collapses this whole notion that every service needs their own. Right. Yeah. Dang. Right. And then, and then there’s just that in the middle of all that too. Right. So 

Kevin Long: outside of Intel, where. There’s like the monolith cloud for that. It’s it is incredibly difficult to do that. And it really does take a bunch of political will and the willingness to. I’m sure it feels like tilting at windmills sometimes to get it done. But yeah, there’s. There’s so much, while it’s not 100 percent overlap, there’s so much commonality around, around some of that, that there really truly is, uh, efficiencies of scale that have yet to be recognized. 

Ashley Nichols: Um, let’s talk a little bit about, uh, the concept of, of full stack development, right? And this has been for the last several years, you know, full stack, full stack, full stack. Is that still the trend? Is there are you seeing a bit of movement away from that back towards specialization? You know, what I read is a little bit of both. And I think it depends on the maturity of your organization, right? If you’ve been in the full stack model for a while, uh, in terms of developers. I think they’re noticing now that there’s a need again for some specialization, but you want to talk a little bit to that?

Kevin Long: Sure. So I smirk every time people say full stack because a full stack developer is not eye shaped, right? It is not, they are not equally deep across the full breadth of their experience. They’re going to be T shaped, right? They can do a lot of stuff. They’re very good at a couple. Um. Uh, and so I think that the, the stated, uh, full stack, I think successful organizations that even while they’re looking for full stack are, uh, hiring people that know how to do a lot of different things, but that they are, have always been paying attention to the specializations within each full stack developer. Um, the really successful ones do make a distinction between like full stack developers and like Product owners and like UI, UX and things like that, uh, where they are truly different things. But the goal of agile is anybody can pick up any ticket and execute it. And so you need to have familiarity across the board of it. But, um, I’m still seeing, I guess, to really answer your question, I’m still seeing people asking for full stack. Um, but, uh, the successful teams that are delivering full stack are paying attention to not, uh, not just that they have experience with, you know, uh, infrastructures code and automated testing and, uh, insert your object oriented program language of choice here and JavaScript and, and things like that. Um, it is. That they look at what folks, um, uh, are, are. 

Ashley Nichols: You’re still looking for a balance of the skill sets across your stack, right? Across it, yes. Yeah, yeah, yeah, yeah. Um, yeah, it’s like a lot, I always think it’s a lot of things. You know, the pendulum swings one way for a long time and then it swings back the other way.

Um, yeah. 

Kevin Long: Everything old is 

Ashley Nichols: new again. I’m, 

Kevin Long: I’m seeing more differentiation on the, not the full stack developer, but on the dev ops side where it’s not, where it was just dev ops engineer, where it’s everything to now you have Kubernetes specialists and things like that, where, where they’re, where they just took, you know, full stack infrastructures code, like people that worked in the server room back in the day. Dating myself now, we’re lumped together, right? You know, say, Hey, I know how to administer Unix and windows, right? Um, but now it’s, uh, I can do the infrastructures code. I can set up the pipelines. I can run that stuff. Or, uh, I’m good at day two ops. I’m good at, you know, the Kubernetes, uh, virtualization control plane stuff and all that. You know? So I’m seeing that actually get more granular than on the day. On the straight development side these days. Yeah, 

Ashley Nichols: absolutely. Um, and we’re going to host another discussion around AI more fully, but, uh, what’d you hear about AI at ProductSpin? 

Kevin Long: Oh, it’s that it’s everywhere, that it needs to be done smart. I’m going to summarize. This is the, the. The Kevin summary of it, uh, AI is only as good or as dangerous as the data and rules that are put around it. And so you have to be incredibly careful with it because it is, it’s ability to operate at speed and scale is such that if it is not thought out carefully and appropriately in the beginning, you will get something that you truly do not want on the end. 

Ashley Nichols: Yeah, I think overall I’m hearing. Excitement than, oh, huge excitement, peer awareness. Uh, but, uh, but yeah, like, you know, what is the governance? What is the governance, um, look like? You know, I heard, uh, at, at a, at a different conference at at ELC, I heard some folks talk about, um. You know, the State folks, the State Department folks and some other folks talk about, um, bots, right, and how they’re going to use those, and, uh, somebody, I will not name them, was immediately sort of like, ultimately, like, we want unattended bots, and then everybody else on the panel went like, ugh, I don’t know, unattended bots? Like, we’re a long way from that, right? Um, but the idea is to, um, take away a lot of the, uh, Human oriented tasks that are like tedious and rote and over and over again. Human checking. Sure. Attended. Yep. Versus just like letting it go. Yeah. Um. 

Kevin Long: it is a supercharger. I 

Ashley Nichols: mean it really is. Yeah. I mean I think we’ve all found that even in like the, in the simplest ways. Right? Being able to use things like chat QPP and whatever to like short circuit, you know, Yeah. Summarizing things that would normally take us, you know, a lot longer and, yeah. I 

Kevin Long: mean, I, I trusted, but verified, I mean, it planned one of the best vacations I’ve ever had. 

Ashley Nichols: There you go. Helpful tip. Use AI to plan your vacation. 

Kevin Long: Absolutely. 

Ashley Nichols: Yeah. It was great. All right. Uh. Well, we’re shifting gears here as we come up towards the end a little bit. Um, what’s the future look like for software factories? We talked a little bit about it before, whether now there, you know, are a few years into software factories and some little searching is going on around the model, um, What’s the future of Software Factories?

Kevin Long: Boy, I wish I knew that answer for sure. Because then I could go, go solve all their problems up front. Um, so it’s interesting. I, I think that they’re still figuring out the way they want to do it. We’re seeing the different services and different agencies approach Software Factories. in a lot of different ways. You know, like Air Force, uh, spins up a software factory around a unique problem set and then they solve it. The Army stands up, uh, like the one ring of software factories that pulls in problem sets from everywhere and operates like a horizontal. Um, you get folks at, at, at DHS that are doing it by Uh, by wide product line, right? And so, um, I think that the future of software factories is going to be more and more around them, figuring out the best way to organize themselves to solve the problems that they’ve been created, to solve the, the power of the software factories and the mentality and the ethos that drives things forward, really? Mm-Hmm. , right? Because otherwise, in my opinion, it’s just. It’s just an agile software team, which is great. You get a lot of things done. Uh, it’s the, the, the scientific method in the questioning and the, the, that they, uh, that software factories apply to their culture and change management that I think really make them. the, the super powerful, uh, change agents for good that they, that they have been. And I think that they’re still looking around and trying to figure out what is the best way, not just necessarily writ large, but for within their agency, for within their problem set, for within that. And I think that we’re going to see them reorganizing and trying to figure out the best way to do that as they, as they, continue to push forward while trying to, I mean, put the plane together while it’s taking off. Right. Cause there’s still, it’s still very early on with that. And so, um, yeah, that’s, that’s what I think is going to be happening with them in the future. 

Ashley Nichols: Yeah. I think we’re going to help them 

Kevin Long: do it. 

Ashley Nichols: I think we’re going to see a lot more, um, focus in the development of the soldier developer. Right. Uh, we, we’ve seen some of that, right? There’s pressure coming from a lot of places, right? There’s too many contractors in the space, you know, what are, what is the future for, uh, our service folks, right? And, and getting into new, you know, areas of expertise, you know, on a lot of the services have the, um, the boot camps and the training to create soldier led software factory. Uh, there’s just criticism about can you effectively spin all those people up from zero to be effective by themselves, right? And have them be 

Kevin Long: effective before they PCS to their next, to their next, uh, uh, posting. Yeah, 

Ashley Nichols: yeah. So I think we’re going to see, I think that there’s a lot of love with that program, obviously, and for good reason. Um. But I think we’re going to see a lot more work on that going forward on how to make that the most effective. What is the best balance between the soldier developer and the, um, the contractor balance, right? Some of the more commercial based expertise bring into the space. So I think we’ll see. I think the more emphasis. On that, uh, more effective training and support for those, you know, those are developers, um, To make the software factories, you know as as optimize them as effective as they can be. 

Speaker 3: Yep 

Ashley Nichols: All right Final question for Rod, as it was, what are your hopes for federal software development in 2024? What are you, what are you looking to see in 2024, hoping to see?

Kevin Long: I’m hoping to see the, so I love the ethos of Software Factories. So, I would like to, I would like to see, Uh, the ethos of Soccer Factories move more writ large, the way that it is, um, the culture of, you know, being able to, to go anywhere that would have, you know, ideas over rank, right? Where you have, if it, if it doesn’t work now, let’s try it and see how it works and move on, accept the good things out of it, throw away the bad things, take the lessons learned and push forward, right? I would like to see that. Implemented at some of the larger. agencies and the at the writ large. And just, I mean, it’s, so it’s, it’s easier to, to shift a smaller culture than it is a larger one. Right. And so I’d like to see, I’d like to see that more and more of, of that experimentation, more and more of that, that, uh, uh, the, the. Ethos and culture, uh, out that I see out of places like Army Software Factory and Kessel Run and Bein and Section 31, and just like, there’s a huge list of them, like, like the, the folks that get there and get excited. I, I, I, I, I, I see physical changes in the people that do that. I’d like to see that all over. I think it’s great. That’s what I, yeah, 

Ashley Nichols: there’s certainly a level of. Sort of enthusiasm, excitement, fun, dare I say, uh, that, that comes with a lot of those environments, uh, 

Kevin Long: paired with incredible delivery. Yeah, 

Ashley Nichols: it is. And I think that there’s a lot to learn from that model, especially in the, um, fail fast and move on. You know, I think what, you know, rightly or wrongly, there’s this, uh, Perception that the US government is sometimes deeply invested in the sunk cost fallacy, which is I have spent this much money, so I better just keep spending until it’s done. There’s stories of systems and, and aircraft, maybe even that never came to fruition that we spent millions and billions on, uh, because we were afraid to walk away and take the lessons that we learned and apply to something new. And, you know, I don’t think that really serves, I don’t think that serves anybody in the end, certainly, you know, not the government folks working on it. The taxpayers, like, like any of it, um, and it’s certainly, and it’s certainly not the model that any successful commercial organization uses. Right. Um, they don’t continue to take money into failing product lines. Not if they’re going to survive, 

Kevin Long: exist for long. Yeah. 

Ashley Nichols: Yeah. But anyway, all right. Well, with that, I will go ahead and wrap it up. Thanks, Kevin, for for joining us. Love it. Um, everyone, thanks for listening to Highlight Cast. Uh, to keep up with Highlight news and activities, follow us on LinkedIn or visit our website at HighlightTech. com. Uh, tune in for our next episode, which will be 2023 in review. All right. Thanks, Kevin. 

Kevin Long: Thanks, Ashley. Have a great day, y’all. 

The views and opinions expressed in this episode are those of the hosts and do not necessarily reflect highlight technologies and or any agency of the U. S. government.