BigBear.ai Holdings, Inc. (BBAI) Earnings Call Transcript & Summary

Hello. Thank you for joining. I'm John Jannarone, Editor and Chief of IPO Edge, with a special event this afternoon. We have 4 guests, 3 of them from BigBear.ai, which, of course, is merging with GigCapital4. We have the CEO, CTO and CFO. And we have the CEO of GigCapital4, who's been a guest with us before. You're going to meet our guests momentarily. Before that, I just want to take care of a couple of quick housekeeping matters. If you'd like to ask a question, and we hope you do, please submit them right there in the Zoom Portal. They'll pop up and we'll get to them towards the end of the hour. You could also shoot an e-mail to [email protected] and we'll find them there as well. And lastly, if you'd like to watch a replay, 3 places you can find that very easily. One is to go to IPO-Edge.com about an hour after the event or just look up the GIG ticker, which is GigCapital4, of course, on Yahoo Finance or on Bloomberg terminals and you can see it pop up there just as easily. One thing we'd like to remind you is that this merger is effective, and the vote will be on December 3. So for any of you who may have had any issues with voting, there's some instructions right here. The easiest thing to do is probably to e-mail [ [email protected] ] and they can help you out. We'll post the same information on the replay article for anyone who's having any issues. But in general, voting is not too hard. You just go to your broker's website and just take a couple of minutes. With that, I'm very happy to introduce a guest who's been on our program multiple times, Dr. Raluca Dinu. Raluca, thanks for joining us.

Thank you so much for having us today, John. Thank you very much.

So Raluca was actually with us last week on our SPAC to the Future event with a couple of other SPAC sponsors and she shared some perspectives on what she's seen through multiple transactions. But Raluca tell us for GigCapital4, you're, of course, the CEO. What were you looking for in this case? And what made BigBear.ai stand out so much to you?

So John, thank you one more time for the invitation, and thank you to everybody that's joining us today to talk about the combination of GigCapital4 and BigBear.ai. A few words about GigCapital4 and GigCapital Global in general, and I'll address your question, John. So GigCapital4 is a $359 million SPAC [ accepted ] in February 2021. GigCapital Global is a serial technologies SPAC issuer. We're working today on our 6th SPAC since 2017. We're a group of Silicon Valley-based technology and global executives. So it's many years of experience in the business. We're all operators. And our mission is to team with bright, ambitious entrepreneurs to provide ongoing support to their fast-growing company in order to become a public enterprise. Through our mentor-investor playbook, we do commit to the company 3 to 5 years long-term partnership and, of course, the combination closing. We're so very excited to be partnering with BigBear.ai team. As I said, we focus on technology and when GigCapital4 became public, our team decided to look into targets in the artificial intelligence machine learning company targets. BigBear is a leading artificial intelligence machine learning company powered by big data and analytics provider. It not only transforms the data into actionable insights, but also provides action-oriented workflows for critical decision-making a real time. Think of this company as a provider of ostensive technologies for key data-driven decision-making problems in large enterprises. Knowing about what happens in the world faster than ever possible has a plethora of multidimensional use cases in the commercial world. We got excited about BigBear because of the technology, which is underpinned by the multiple intelligence community agencies that repeatedly reach out to the company for some of their most complex challenges. We got excited about BigBear because they have best-in-class leadership with world-class academic and industrial pedigree. We got excited by BigBear.ai because it is a company that's coming from 4 M&As and in itself is a very solid platform for additional M&As to grow inorganically on top of very strong organic growth. John, what do you think about that?

I think that's great. All right. Well, you know what, let's bring on the rest of the team here. This is very exciting to have all 3 of you from the C-suite. So thank you, but I'm going to let Jarrett take over for a moment. So Jarrett, why don't you introduce the group here and kick it off for us?

Thanks, John. And it's a pleasure to have Brian, the CTO, Josh, our CFO; and of course; Reggie, the CEO. Reggie, I guess, we'll start with you. Now this company has actually been around and have the track record of 20 years, which some people may not know. Can you tell us a little bit about the foundations of BigBear.ai?

Absolutely. I appreciate that. And thank you all for having us. Great to talk with you. So I think I'd like to go and get a little personal about this, okay? Because coming out of government where I had background in DHS and DOD and DARPA as well, and one of the things I knew that was important and for the government in particular, when I first was looking at this is artificial intelligence. And there's a report that was a co-authored by Bob Work, who is former Deputy Secretary of Defense, and Eric Schmidt, obviously, of Google, who talked about the importance of AI for national security and national defense. And this is something that has really struck me as an essential part of defense intelligence moving forward. So coming out when I was approached by [ Eight Investor Partners ], that's the private equity firm that [ the project went ] together, one of the things that I was really interested in was this AI kind of platform. But it wasn't just AI in general. It was a particular type of AI that we're talking about, right? And that's when you get to the problem that we're trying to solve. So what's that problem? The problem is decisions. And why are decisions a problem? If you listen to the Chief of Staff of the Air Force, he talks about imposing multiple dilemmas from multiple domains and an [ adversary ], right? That's the kind of environment [ that warrants ] national security. If you look at the business flow, there are these -- this is a complex environment of decisions. Consider what's going on right now in supply chain. That's what's going on. You've got manufacturing at the manufacturing components. You've got the labor, the supply of labor, you've got the transportation issues. So you've got this plethora of issues that impact the supply chain, and we're going to make decisions. You've got to navigate your way through this complex and sometimes chaotic decision space. So the problem we're trying to solve is how do you navigate through that decision space, and that's what we do. We provide the forecasting. We provide the rapid assessment of course of action as you are making these decisions, [ one of them ] happens. And what we pride ourselves on is giving advice. We give you advice. So the history of BigBear, it comes from the problem we're trying to solve. And that's the navigation through this complex and complicated and chaotic decision space, right? So what do you need? You need analytics, you need the data, you need ways to present this information. And given the world we're in, you need the cybersecurity framework, right, around -- to frame all this. So that's what -- Luca talked about 4 companies coming together to build BigBear.ai. These are 4 companies that have histories. They're histories of 20 years of profitable performance. And they were brought together because of their analytics, because of their data, because of their visualization because of their cybersecurity capabilities to form what is now BigBear.ai. It now provides the kind of navigation to this complex decision space that we do right now. And these 4 companies were all profitable over their entire histories and growing. So that's something we really bring that is unique. It's not just the capabilities, it's not just the profitability. There's something that I left as well. One of the other important things here as well is innovation. And so when we're talking about companies and talking about bringing companies into this platform, it's not just about the forecasting or the data, but it's also about how do we innovate. Because I think in order to actually solve problems, real relevant important critical problems, it's essential to differentiate between invention and innovation. And I would argue that invention really has to do more with a cool idea, right? Where innovation is a cool idea that solves a real problem. But under -- to solve a real problem you have to understand what a customer's challenges really are, how they work, their workforce, all of those things. So one of the things that we found with the 4 companies that came together to make BigBear is a true innovation culture. I think that's important. One last thing and then I'll be quiet here, one last thing is when I was in DHS, I had this really interesting experience. I had a chance to talk to a bunch of CIOs from some large financial institutions. And one of them shared with me is -- it led to a cybersecurity discussion, and what we were talking about is what are your biggest problems? And he said, my biggest problem -- and this was surprising to me -- is integrating tools into my existing infrastructure. It wasn't -- I need some new tool, it was the integration problem, right? So that's the fourth thing that we talk about with BigBear, and Brian can go into a lot more detail about this is how we seamlessly, frictionlessly integrate into the existing infrastructure. Talk to Brian about that.

That's great. Brian, maybe you can tell us a little bit about some of the services and advantages you have over competitors?

Well, I appreciate you throwing it to me. I'll actually start by mentioning that I am in San Diego right now at the Point Loma Naval Base, hanging out with the naval warfare folks. Not 2 hours ago, we just received first prize in their AI advanced naval technology exercise, came with a $75,000 price. We beat out Booz Allen and Northrop Grumman and Lockheed Martin, L3 Harris, a whole bunch of other competitors. Our command and control AI that we're using here to modernize the way naval commanders are making decisions strategically and tactically is already just another example of how we transformed the decision space that our customers operate in, right? So they're loaning me this room that I'm sitting in, I'm bothering the guys here, but the admiral, the rear admiral put us down here. And honestly, we're super excited to be continuing our growth in the federal space. But I think your question was what is it that we do that differentiates, right? How do we describe what we're bringing to market? And I like to describe it as -- we are all familiar with the way we drive today, right? When we're navigating from point A to point B. We all know that we pull out our phones and we bring up our favorite mapping application, we type in our address and it routes us. And we have grown to trust that AI helping us make that decision. And you think about the path that we went on to get there, right? We started with people driving with a map of the seat next to them, kind of looking at the traffic right in front of their noses, trying to make decisions with limited information. And honestly, that was -- we all grew up with that. it was not 15 years ago, that was the avant-garde. And then the radio stations said, "Hey, we can we can help with that. We can start reporting on traffic conditions around certain well-trafficked routes. And the drivers of those routes, you can listen to the radio and if you're lucky, you can hear about an accident, make a decision before you run into that traffic, right? But now with the remote sensing that the world has brought us, we have telemetry on cars. We have overhead imagery doing collection. We have so much remote sensing that we've been able to build an AI that can route drivers far better than any driver could do on their own. What BigBear is going to do is allow that same sort of navigation support for your enterprise decision-making. If you come to our platform and you tell us, we'll show you where you are. Now you pour your data into our platform, where you use the data we are already bringing to bear. We show you where you are now, you tell us where you want to go, and we will define, we will identify the routes that take you from point A to point B. And that's not just a spatial track through the road network, that is budget allocations, fleet assortments. It is all of the domains that impact your business. You might be moving levers, pulling levers in those different domains in order to make that much more complicated decision. And our products are there giving you data-driven advice to what those course of actions should be. And if you take a certain course of action, what the likely impacts are. And not just immediate impacts. We're talking about second, third, fourth order impacts because to every action there is an equal and opposite reaction. So our machine learning, the tensor completion that we specialize in is built to do that. And honestly, I think that is going to be so transformative. It is going to be the must-have enterprise decision-making product for the next 20 years, 50 years. That's really going to transform the way people operate in the future.

That's a pretty exciting stuff. Now gentlemen, you've described Big Bear as providing a bunch of LEGOs. Can you tell us what you mean by that?

Yes. The -- one of the things that we -- we've been in business again for 20 years, and we're working in the government where the government obviously has some massive investments in other technologies, right? And it's unrealistic for us to walk into any customer and say, "Hey, stop what you're doing. You've got to start all over with BigBear. Start to finish, we're changing the way you do things, right? And we can do that. We have -- that's actually why, as Reggie was mentioning, we brought together these 4 businesses because now we can offer an end-to-end platform from data to decision, we can do everything in-between. But we're not forcing our customers to do that. We want to say, listen, if you've got your favorite data layer, if you've got a user interface that your senior leaders are already conditioned to leveraging. If you have your own data science team and you've built your own forecasting models, we can plug all of those things into our platform because everything we've done has been built as LEGOs as you said, right? Every little micro task, microservice that we've built is wrapped in APIs and deployed onto this common event-driven bus, so that we can snap those LEGOs together in any order that our customers need in order to meet their purposes. So that makes us very easy to work with. It makes us so that we can provide value very quickly for our customers. It also allows to specialize in the things that we do better than others. And so we don't always have to be the best at everything. We can really focus on what makes us different in the environments.

Can I interrupt for a quick second because I think Brian mentioned something. He made this phrase, data-to-decisions, right? But I think we'd like to talk about dirty data to decisions, right? And so Brian, can you talk to why we say dirty data decisions or why that differentiates us?

Yes. That absolutely is a huge differential. And again, it comes from the fact that we are -- we were born in the defense space, in the intelligence space, where our customers today, but know we have an awful lot of federal customers. And those customers don't always own the data that they want us to process, right? They're kicking in doors and getting little snippets here and there. They are -- in the intelligence community, we might be getting that data from other means, right? So we don't often have perfect data. We have sporadically collected data. We have data that our adversaries are purposefully spoofing so that they try to give us wrong data, right? We have all these data problems that our defense and intelligence customers need to be able to deal with in an automated, robust way. And so because we've invented our machine learning to deal with that dirtiness, honestly, we found that dirty data is pervasive in the real world. No company has perfect data for every domain that they want. Think of it, just competitive intel. If you're trying to understand what your adversaries are up to, it's not like they're going to give you their sales book, their point-of-sale transactions and say, "Hey, here you go. Go analyze our business." That's the kind of thing that we can do for you. We can fill in those gaps in your environment using the contextual sensemaking that our tensor completion is capable of doing. And that is a huge advantage for our methods over our competitors. Our competitors tend to start in retail, where your data is [ nigh on perfect ]. You know with every transaction you do, you know every product you sell, you know every customer that walks through your doors. You might be missing some competitive intelligence, but when you start working into domains like manufacturing, where you don't have visibility of the full supply chain. You might be working in site selection, where you don't know what all the construction companies are planning on doing for the next 5, 10 years right? In those domains, you have imperfect data, and we have the methods to let you deal with that.

Great. I'm going to hand it over to John here. He's got a question.

That's great. Let's bring Josh into the conversation. Josh, thanks for being here. I think that what I'm going to ask relates to some things that have already come up. There's a 20-year track record. The company is going to be cash flow positive this year. But something that I think is important in the context of this SPAC environment is to talk about what underpins the projections the company has laid out. Can you tell us a little bit about the pipeline or ongoing business that's actually supporting those and why it shouldn't be viewed as speculative?

Yes, absolutely. Great question, John. And that's honestly one of the biggest differentiators. Investors really want to see predictability out there. And this more so than ever, especially with companies going public, either via direct listing IPO or certainly through a SPAC transaction and many of them being pre-revenue companies. So as you are well aware, we have $0.5 billion of backlog already. So this company has been performing for decades. As a matter of fact, just in Q3, we won a major contract. This is $140 million that was added to our backlog. So this gives you substantial security or predictability of the revenues moving forward. When you look at 2022, 60% of our revenue is already in backlog for the company. So a lot of that is in multiyear engagements that honestly stretch out with our government customers up to 5 years. So we have revenue already on the books that you can see out to 2026. So that gives investors the insight on both the revenue and the gross margins looking that far out into the future. So with that said, on top of that, we have 100% win rate on our recompetes, meaning we haven't lost a customer to date. And on top of that, in terms of our pursuit of new work, we have a 93% win rate. So that provides a great stable operating platform for us to keep building the business moving forward.

Great. And you touched on this, but if I can just ask you to flesh it out in a little more detail. What does the typical client contract look like? You mentioned some of the government ones are quite long, they're several years. Is that usually how it is rather than a spot assignment?

Yes. So on the government side, we generally see them coming to market, asking for a long-term solution. Most of the time, those are 5-year engagements that we see there. In the case of these government implementations, a lot of times, we see very specialized on-prem solutions. And as you can imagine, with our major customers being the Department of Defense and the intel community, very unique secure domains that we need to deploy into. The other part of those deployments that is interesting is, as you can imagine, they are operating in the most complex data environments. When you talk about the magnitude and the complexity of the data that they work with. So because of that, their requirements are always evolving. And what that means is we're able to deploy into their environments. But then we are continually looking to add new algorithms, new data sources, new analytics and researching those alongside our government customers. So a lot of times, our data scientists and software engineers are sitting right there, running the software for them in some cases, but evolving the technology. And then that allows us to take some of those most advanced approaches and bring it over to our commercial customers. So on the commercial side, completely SaaS-based subscription model. So you see the 1-year recurring subscription is a general deployment there. So for the most part, those are completely hands-off deployments. As Brian mentioned, we can deploy it into their environment that they can operate the software on their own from that point. So the one other thing I wanted to mention on this thread, John, is for those commercial implementations, we have 3 capabilities: our observe, orient and dominate. And we allow customers to buy into any 1 or all 3 of those. And what we initiated or the way we've approached this is allow a customer to come in and try one of the technologies and realize how impactful it is on their operations. One of the most surprising things for us over the last year is while we kind of assumed that would be way it is deployed into a customer space and they would build up over time. We have realized once we're able to demonstrate how the technology fits into their decision-making process, more often than not, customers are not saying, "Hey, I'll try one and then buy up over time." They're actually saying, "I want to see the full suite deployed immediately." So it's been a really positive trend for us over the last year to see how many commercial customers say, "Just give me the whole lot of kit and kaboodle."

That's great. I want to shift the question back over to Reggie and Brian. Brian did a great job earlier of explaining some of this. But just help me out if you can. I'm trying to imagine what it looked like inside of the Pentagon or somewhere in the intelligence community going back a couple of decades. I'm just imagining these reams of paper with all this data before you had any ability to crunch the numbers. How has that evolved? And even now, do you -- does your AI help you sort out and organize which data is most important to look at first?

Yes, I can jump in on that. So you're -- sorry, I mean if you think about it, the human analysis is still the most trusted advice that comes out of our intelligence and defense customers, right? And so that -- but you can't -- much for the same reason that BigBear is taking a product direction. We're trying to take all the intelligence that's in our heads and build it into a product so we can scale faster than we can hire people, right? The analyst community -- the intel community, the analysts are having the same sort of problem where the amount of data that is dropping onto their desks is increasing at an exponential rate day by day by day. And these analysts, they've got the point where even if they're super specialized, one just shining a flashlight on just one part of the world. They don't have enough hours in the day, even to go through all of the data that's relative to that one spot. Let alone watch the rest of the surface of the globe for other things that might be happening. So over the past decade, really as we've been involved, we've been able to produce a triage mechanism. Think of it that way, right? We have this observation layer that is constantly able to look at all of these different data sources across the entire surface of the planet and conflate those sources together to start making observations about behaviors and entities and activities. And those -- it's kind of a base level of triage, a very pervasive, low-level observation that then provides tipping and queuing, alerting that goes up to the -- as we find something, as our AI says, "Oh, something's happening in Sudan and that is probably going to result in some kind of effect that someone needs to know about." And we start learning people to say, "Hey, we need human eyeballs on this part of the world or we need a better sensor data about this part of the world so we can do a better automated analysis job." And so this triaged mechanism is what over the past 10 years, has become a huge enabler of our analyst workforce because now they're not trying to split their attention 1,000 ways. They're able to focus as much as they can on the problems that we bring them to say, "Hey, here's where you need to focus right now. This is the most impactful thing that's happening in the environment that's going to result in these outcomes and you need to study it to make sure that you understand that we're -- either confirm that we're right or tell us we're wrong, but you need to pay attention here." And that ability is -- think of it as we're able to focus attention. I mean, everyone has heard about the attention deficit economy, right? Attention is a finite resource. So we going to help our analysts, we've got to tell them where to apply their attention so that they can make the hard decisions with those human eyeballs because they can't keep pace with everything that's falling on the desks. So you're right. It was literally just piles and piles of paper, just reams and reams and you had these rooms with massive printers printing out maps of the world, right? And honestly, all that's gone away now to the digital workflows that we are now enabling. We're now the only [ imprinted ] AI actually in this -- for one of our particular intel customers, that is producing what's called actionable tasking. And that means when our AI identifies, this event is happening in that place, that is actually able to task analysts in the field to go and do something about it, right? So we are alone in doing it, but the point is that has made them so much more efficient that the director of that agency has now given us 2 different awards for transforming the business process and letting them stop trying to just hire, hire, hire and triple the size of their workforce to keep up, but be able to do things smarter and be more agile at the same time.

Let me add something to that because I think what's going to be really interesting has been -- not just on the federal side. But as we've got inquiries from the commercial side, right, we found that are in a similar situation with their data, and we could [ never ] go in and help them understand not just what their data shows them, but the other things they could do they had no idea they could do, right? No idea. And this has been true, even we've been approached by restaurant businesses. Brian mentioned site selection, right? We've been approached [ 5 floors ]. And one of the areas that we're particularly excited about right now really is these commercial space opportunities. And the fact that we've announced our partnership with Virgin Orbit, Redware Space. And so we are really excited about the opportunity to be analytics provider for this really growing -- this growing industry around lower orbiting satellites, right? [ We're around ] how large that is, what -- they're providing Internet services, whether they're providing different types of imagery. This is a space where we can no kidding, make a huge difference. And we're able to show the companies, what they can do with that they had no idea they could do. They confuse this information. They had all kinds of other insights that they didn't realize because one thing that going to talk about is where we conflate data, right? And what we mean by that is we're taking information in different types of sources, putting it together, given a certain geographic location and exposing what that larger context means, right? So it's not just this is a supermarket. It's what's going on in that scene. And so by taking information from a lot of different types of data sources, we're going to conflate that and then provide a lot of insight of what that [ even ] means. This is the kind of insights that we're giving to potential customers that they had no idea they can gain. So I think, while you've got these traditional scenes of people with reams and reams of data, you've got people with data right now. They don't even know what they -- if they knew what -- they don't even know what to do with it right now. And we're able to show them what they can do with it with our expertise and with our core capabilities.

That's really helpful. I think we might want to do a little more into the commercial client opportunity in more detail. But before that, there's a slide in your presentation that I really liked who showed a few examples around the world of what your data allowed some of these big government clients to do. There are a couple of examples there with Libya, Crimea. Reggie, we don't have to go through all 3, but I think it's useful for the audience to see what the actual real-world impact of your work can do?

Yes. I appreciate the question. So if you remember, the Russians invaded Crimea. And because of our ability to conflate this informational variety from different types of sources, because of our ability to train an AI based on similar types of happenings, events, we're able to predict the Russians invading Crimea long before it happened. That's significant, right? That is very significant. Another event you mentioned was the fact that we know that -- we found that vessels that were actually flying a flag in of police vessels we're actually [ Libyan ] vessels. And this resulted in actually the arrest by the coast guard of these vessels, these were doing illegal and illicit activities. I could also mention the Iranian activity, we're actually there to help U.S. Central Command with their Iranian strategy because the insights that we can provide for the senior leadership. So I think these are some real-world examples and there are others on the kinds of impact that we can make with our assessment of the situation and our ability to forecast.

And by the way, I should point out to anyone who hasn't seen that presentation, it's easy enough to find. Just Google BigBear.ai investor presentation, and it will pop right on. Jarrett, I'm going to hand back to you. I think you're going to ask a little bit about connected devices and how that's changing the world and then the kind of data that's out there.

That's right. Brian and Reggie, a question here for you. It's a large and growing total addressable market driven by connected devices, right?

Yes. Absolutely. I think that when you look at [indiscernible] defense, the [ IRT ] if you start -- and the tremendous amount of data. If you talk to data scientists, they talk about the volume, the velocity, the variety of the data, right? So when you talk about the [ IRT ], you get all of those because now you've got individual sensors, potentially billions of these sensors, cameras, doorbells and everything else, all computers, all connected. And these are all producing massive amounts of data, which then can be used to do the kinds of things that we mentioned that we do, [ which is the place ] your data plan in just they have the completion date, the understanding of the situation, whether the forecasting and the advice you give them, right? And then -- so you talk about that, but then you also get into the other area that I mentioned, which is these commercial satellite consolidations, which are also provided, remote sensing information and others and you start seeing just this proliferation of data. That again, it's massively increasing the volume, massively increasing in the velocity, the speed at which we're getting this data and massively increasing the types of data or the variety of it.

And it's not just the government contracts. You guys are also seeing a growing -- an emerging opportunity in the commercial space, right?

Right. Exactly. So that's what's getting on an earlier -- in particular areas, right? While we don't see ourselves as a platform, we're not going to solve all problems ofr everyone. We're going to solve specific use case problems. And we're working on right now is [ space ] so just -- to just repeat. This is the problem if you've got lower orbit constellation, [ sale ] constellations, there's a variety of types of sensors and [indiscernible]. It can be synthetic-aperature RADAR, it could be LIDAR, it could be hyperspectral imagery, Electrooptics, IR, these kinds of things. And we [ confuse ] that information. We can then make sense of that information presented in a way that someone can actually know what kind of decisions to make based on what they're seeing at a given scene. And this is all based on the kinds of things we do right now for our current defense customers. So it's easily portable to commercial world. And one thing that Brian has done a great job of is taking the kinds of things that we do for our defense customers and say, what can easily map over the commercial space, right? What are our areas of strategic expertise that can easily map, and what we're finding is that we're getting a lot of incoming from commercial customers saying, "Hey, can you help me out?" And we're finding is, yes, we can. Brian, do you want to add to that?

You definitely captured it well. I want to throw it right back over to Jarrett, but I think we've got a pile up of questions. I want to make sure we're able to get through.

Yes. Well, well, first, though, Jarrett, I think we're going to talk about Palantir, which of course, went public recently. So when we spoke with the group yesterday, I asked, is this a competitor, but it turns out it's a little bit more complicated than that. So I don't know if Reggie or Brian wants to jump into that one?

So let me start and then I'll throw it over to Brian. I think it's a strategic partner right? I mean I think in conversations with Palantir, it's become clear that they have a platform offering, a scaffolding, a framework in which we can plug our differentiated AI solution. And I think as we go to market like that, I think we bring significant benefits to both companies and to our customers. And Brian, you can give some more detail.

Yes. I actually think you were asking us about our LEGOs earlier in the conversation. I think our relationship with Palantir is proof that our architecture, the way we have these LEGOs is incredibly compelling because we can plug those LEGOs into other people's platforms to make their platforms better. Palantir has recognized this. They recognize that what they're providing is this general purpose operating system for the enterprise of tomorrow, but it's not verticalized. If you want to address a shipping logistics decision or a retail site selection decision, they don't give you an out-of-the-box capability. But if you were to plug in the data sources that BigBear brings to bear, the analytics and the simulations that BigBear can plug in, then suddenly, they have an out-of-the-box capability that does directly meet those business needs and it makes them able to sell more software. It's a great channel partner for us so that we can sell on top of their installed base. And it also allows us to really even spend more time specializing in the things where we do things differently, better than others on the market, right? We don't want to reinvent the wheel. If there's things that we can plug into where customers have already made those investments, let's not try to invent a better mouse trap. Let's focus on continuing to do what we do better than the market does. So we are very excited about the Palantir relationship. And honestly, we're looking for other places where we can have channel partners that will be similarly helpful to our go-to-market.

Can you talk to us about this idea of becoming a product company and still attracting top talent?

Absolutely. Right. I mean I think -- so first of all, obviously, tracking top talent is different, it's different for everybody. It's a challenge, particularly in this field that everybody has. We have a 90% retention ratio and the reason why we have that retention ratio is because we provide a number of different opportunities to our folks, right? Prior to the opportunity to work on just cutting-edge problems. And these high-end data scientists, engineers and architects, that's the kind of thing that they're looking for. But we also offer something else. We offer -- I mentioned this in kind of the beginning which we also offer a culture, which enables them to really get to understand the challenges, the customers better than even the customer understands. We give them a chance to solve critical problems in the business world, critical problems in our security world and where we trying to -- where we're now expanding into is in the commercial product space. The commercial product space, now we have these SaaS offerings, they enable us to scale a lot more rapidly than just individual solutions with a set number of engineers and technical folks. So going to the SaaS model allows us to scale much more rapidly than we could just going with a team of folks going out given the unique -- given the unique custom bespoke kind of solutions.

Great. We've got several questions here. But before that, John, do you want to ask...

Yes. Yes. I do. I think that something that jumped out at me is that you, in fact, already have an M&A pipeline, that's part of the planned use of proceeds. Reggie, can you talk to us a little bit without getting -- I know you work in a very secretive world, without giving too much away, about who those potential targets are?

Yes, sure. Let me give some general framework, right? It wouldn't be any surprise, right? So we talked earlier about the kinds of things that build the current platform. Its data, its differentiated analytics. It's visualization. It's full spectrum of cybersecurity. So these are all areas that when we are still interested in building our capabilities, as well as building in adjacent markets right? So we can talk to that. And we've got a pipeline right now that we're talking about in our briefing. Another 25 companies that we're looking at right now. And we expect after this, that the transaction closes, that do we even increase M&A? Because I think one of the things that we think about often is this build versus buy decision, right? So as we start thinking about increasing our capabilities, increasing our market penetration, increasing our customers set, do we build or do we buy? And a lot has to do with timing of the market. So when we find a company that can help us in those areas, that's something that we are going to want to acquire. And we plan to be quite acquisitive going forward and looking forward to opportunities to talk to various companies in there.

Great. Jarrett, do you want to take a couple of questions that are coming from the audience?

Sure. And a reminder to the audience, please keep them coming. We love them. So here's the first one, do you compete with scale on AI? And what are the closest publicly traded comps you compete against?

I'll jump in on that one first, Reggie. So Scaling AI is not a competitor. We actually see them as a partner. We've met -- they are -- honestly, they are a great vendor of label training sets and data that we can take advantage of right -- I think that, that is not someone that we've ever competed with, more often, we would be inviting them to join us in our deliveries for different customers. Now talking about comparables, who do we compete with? We talked a little bit about Palantir. I think one of the reasons we're looking to get close to Palantir is that we find ourselves in the same circle so often, and we're both specializing in very kind of different synergistic places in those competitions. So why not merge the 2 best offerings and really knock those competitions out of the park. So I think we see a lot in the government space where we've been around the government so long that the government sees us as a platform, right? They see us as their analytics solution for everything. And so we tend to compete more with the platform providers in the government. Those would be the C3 AIs, the Databrick's, et cetera. In the commercial space, we're taking a much more vertical SaaS approach. We are offering vertical solutions that solve very specific business problems. And so for instance, in our media analytics offering, we compete more with the data miners of the world, right? Because that media analytics is an established vertical, and there's tools in that vertical. So that's really where we're focused on the commercial side. And we really don't see a lot of the platform competition affecting us in the commercials. And actually one of the reasons why the Palantir partnership so great because we can actually ride a lot of their platform deployments and allow us to accelerate verbalization.

That's a great answer. We've got another one here. Please expound on the subject of the integration problem mentioned earlier, data sources and BigBear approach?

Yes, absolutely. So the challenge really there is you go to a number of particularly -- these institutions I was talking to. They have a lot of equipment infrastructure already on track. And the question then is, if I'm bringing in some new system, how do I integrate this thing without tearing up into a forklift operation and what they have right there. So that is with the data, it's with the tool sets, all of these things become a problem. And because of the API center architecture that we have and the composability. Actually, we even talk about that, the composability of our platforms, we are able to integrate more effectively than others. One of the things we talk about is the fact that we can integrate best-in-class attributes of other people's stuff, right? And I think that's an attribute we have as well because sometimes if you go into one of these companies, they don't want to just tear up what they have to put in a complete end-to-end system. They want to embrace something that has synergies, that helps and improves their existing capability. Because of the way we're architected, we're able to do that better than other people can do that. It's because the composability that we have, that we talked about the LEGO bricks before, the LEGO blocks? It's that LEGO blocks concept that allows us to solve this problem.

That's great. I think this one is for Josh. What is the appropriate level of cash on the balance sheet to execute on your business plan?

Jarrett, good question there. So one of the things we're most proud of is the company has been cash flow positive for years. So for that reason, the investments we're making today in the workforce and the R&D moving forward is completely internally funded or can be. And so the cash to the balance sheet, while important in the transaction is really not for funding current operations or what we're projecting today, but an accelerant to move into these new markets. So it's not really a requirement for us or an important consideration as much as a catalyst for how quickly we can move into some of the markets that Reggie spoke to.

If I may add to this. As we are getting closer to our shareholder Board meeting on December 3, we already raised $200 million in convertible notes and GigCapital4 has about $359 million in the trust. Additionally, BigBear.ai is supported by one of the largest [ P ] in the country, that is an [ AE ] Investor Partners, a very close partner of GigCapital team that supported the company to date through the M&As and we're certainly supporting the company into the future as the team is becoming up of the company. So we will certainly support the company and the balance sheet that Josh has talked about and Reggie and Brian referred into the public life.

Sure. Let me jump in with one related to M&A. Reggie, you touched on this a little bit, but there's something that's in the presentation deck. And by the way, a couple of people just ask where to find that, I posted that in the chat there, but I'll tell you we'll put it in the replay article as well so everyone can find it, the link that is. But there's a mention there of moving into -- moving out horizontally into different areas where you're not right at the moment. Can you give us a glimpse of what those might be? And I don't want to emphasize this too much because that might be a little further ahead than what you're focused on right now, but I think it's still interesting.

That's fair. So again, when we talk about the kind of skill sets we have, location intelligence, maritime intelligence, media intelligence. These are 3 areas that we're specializing and developed this specialization as I mentioned before [ work in gardener work ]. So these are the areas that we try and we can then move into horizontal categories. Some of these horizontal areas that I think mentioned in the deck are healthcare, finance, these kind of areas. And we're actively working those kind of use cases. But I think one of the things that we need to be clear about is we are very focused on having the right use case, the right vertical market to go in, in these particular areas that is really -- that our capabilities are uniquely suited for. We're not trying to solve all across for everybody. So as we move into these particular areas, that will be backed up by real assessment of the market, a real assessment of our capabilities and real assessment of the challenges that these customers have.

All right. Great. I've got one more then I'll pass it back to Jarrett. There's another good slide that I like in here, which explains the rationale for the peer group that you chose. Can you talk a little bit about who the best comps are? And I mean I'm not sure who wants to answer this, maybe that's for Josh, if we're talking about how you think about valuation or it could be for Reggie. I mean if anyone goes to look in there, you can see that the company is trading right where it is a little bit around $10 a share. Is it a very steep discount to a lot of these competitors. So it looks price with plenty of room to run. But why did you choose those and which ones should investors focus on?

Brian, I was going to say maybe before I jump into a couple of the numbers, you want to talk on the technology side, maybe in terms of capabilities and then I can hit the numbers?

So Josh, I think you had to step away for a moment. We had actually covered that. So jumping to the number I think is actually best.

Okay. Yes. Sorry about that. So as we look across here, I mean, we found these companies that are largely in the AI, machine learning, SaaS-based models. And what we found is we chose these companies because they're close to us in capabilities. And as we look at how we measure up against them, we found that our growth rate even historically, puts us in the middle of the peer group for what these companies were able to achieve. Now with the acceleration with M&A moving forward, we think we'll be above most of the companies in that space. The other part is a lot of these companies are not cash flow positive today, not EBITDA positive. And they are funding this growth with aggressive spends. While we do have an M&A pipeline out there, the historical performance of the company already puts us above most all of these competitors. So when you look at Alteryx, C3 AI, Palantir, Snowflake and such, certainly, companies of different size. But when you look at how we're performing on these revenue and EBITDA metrics, you see how favorably we compare to these using the Rule of 40, which is a combination of that growth in EBITDA, you see that we're leaps and bounds above most of these companies today.

Great. Jarrett, I'll pass it back to you in a moment, but something else I wanted to ask about is this shift towards more of your revenue coming from analytics. Does that mean that you're going to have higher margin revenue over time as you shift? And I think by 2025, it's going to be really -- I mean, it's going to be way up there. It's roughly 50-50 now, and it's going to be much, much higher -- much, much more shifted towards analytics?

Yes. That's a great point and I kind of alluded to this earlier. With many of our historical implementations being with government customers and those being such unique domains that we have to deploy into. There is a lot of customization considerations there, and that's a higher, I'll say, labor components for those deployments and the ongoing support in research and development. So on the commercial side, the gross margins are considerably higher to the tune of 20%, 30% higher than what we see on the government side. So as the commercial revenue ramps up, you see the gross margins for the company improving as well.

Okay. We have a question on your renewal rates and on how many customers.

Yes. So we are deploying our capabilities through 75 different engagements today. And when you think about engagement with our current customer base, you can think in terms of the Army or Defense Intelligence Agency. Those are pretty big customer sets, right? So 75 different engagements, but those can range anywhere from $100,000 to $200,000 of recurring revenue each year up through tens of millions of dollars. Probably the better way to think about it is there's over 100,000 users just in the government space today, using 1 or all 3 of our capabilities. So we see that expanding, obviously, more on the commercial side moving forward. But when we talk about that recompete win rate, it's on those existing vehicles supporting government agencies.

There's something we can add on that, the win rates. So You mentioned that we have a win rate on renewals on [indiscernible] and 93% on new business. What we didn't mention is that when we factor in our pipeline we assume a much more conservative set of numbers. So for our pipeline, we factor 70% for recompetes and 40% for new business. So that's part of our conservative take on our pipeline going forward, which gives us confidence in our projections for next year because of the backlog that Josh mentioned 60%, then the fact that we are very conservatively factoring our win-wins going forward.

Great. Somebody is asking, is Innodata a partner?

Yes. So I'm seriously Googling Innodata right now. We have not worked with Innodata. We've encountered them. It seems like someone we ought to be talking with. So we appreciate the tip and we'll be digging in.

Okay. Can you comment on lockup arrangements?

So Jarrett, a standard lockup agreement, 1 year or the date at which the stock trades $12 per share, 20 out of 30 days consecutive days of trading, starting at 90 days after the moment of the combination. Thanks for the question.

Reggie, I want to shift back to this discussion of all action going on in outer space. We've actually hosted a number of these companies here. I mean it occurs to me more of that data that's being collected out there. I mean if you're able to work with them in a way, is BigBear an investment in the data that's coming from space? I imagine there's going to be more and more of it coming. I mean is it a significant -- is it a seriously significant opportunity for the company?

We see that as very much true, so I appreciate that question. There's no slowdown in the number of companies trying to put constellations out, right? And we're seeing that increasing. I mentioned quarterly, sensing and different types of sensing, whether you've got individual constellations with some type of aperture, different ones with hyperspectral different lens with a -- [ find it on the ] Internet, et cetera, et cetera. But from optics -- and what they typically have, they may have launch capability, right? They may or may not have a ground segment capability as well, but they can launch them and they have the telemetry. What we're providing them is this analysis in the sensemaking of the data that's coming from the sensors as well as providing ground segment support. So that's -- we didn't talk about that, right? So a big part of the -- when you're ready to sell that constellation for the space segment or ground segment. Well, many times, they are putting up these satellites. And if we have a sophisticated ground segment and we have been asked to help with the ground segment as well as the data fusion as well as the data analysis. So I think this provides us a great opportunity going forward between all these different systems that are competing for views [ like ground 4 ].

Now something that I think we may have touched on yesterday and not today is that you need people to have security clearance because of who your big -- couple of your big clients are. Now when you're out getting new business, can that ever any tricky situations that come about? I mean, certainly, I'd imagine there are countries that our government wouldn't want you to work with. But for the most part, particularly in the commercial sector, are you free to go out and seek new clients pretty freely?

So recently, you may have seen the notice that we hired a President of Commercial and President of Fedral. We did that with a recognition that we need special expertise on commercial. And so we are building up a professional sales and marketing team on the commercial side, with that head of expertise and making sure we have the right types of segmentation, so we don't have the kind of problems that you're talking about. But we're very aware of those kinds of issues, given our defense background, and we're very cautious and concerned about how we're approaching that, but that's exactly why we have a commercial segment and a commercial present from on the commercial world, they were hiring into -- we're building that up aggressively right now.

Great. Raluca, you talked for a minute about the question that Jarrett asked. But if I can, I think some investors might not be accustomed to seeing a convertible associated with one of these transactions, pipes are well, it's also dynamic. If you go back not too long ago, there weren't even really pipes happening. But can you tell us if investors should be thinking hard about what that means? Or is there not that much of a difference? I mean, I'm imagining if the convertible is converted, it's quite similar to the impact that they're being pipe, right?

Right, John, and thank you so much for bringing that up. You're right. We have a $200 million convertible note that we will fund BigBear as it becomes public. And it does look like an equity event to a certain extent, it has an interest that the team at BigBear is ready to pay in time. But we do feel very comfortable that part of the trust money will become the equity on the other side in the public market and the company will be properly supported, not needing further cash as it becomes public.

Great. I think in his opening remarks, Reggie talked about the supply chain crisis. It's in the news every day. Does all the data you have and your ability to process it give you any window into how it might be solved any more quickly? Is that something that you guys could help us out with?

Yes. So it's funny, we -- absolutely, we're knee-deep in that. Although our commercial presence isn't significant enough that we're making changes today. But on the government side, we are very much in the middle helping them correct their supply issues and their fueling issues of equipment and the movement of even material and people around the world. On the commercial side though, that is -- literally, that's what our Maritime Intelligence solution is built to do. It predicts congestion. It allows you to run some courses of action to say, well, if all -- if I'm seeing a multi-day backup at the Port of Savannah because they don't have enough cranes to service both their inventory of containers, that's sitting in their warehouse that they need to put on trucks as well as offload the containers that are coming off ships. Being able to say, hey, if I would normally send my vessels to Savannah, should I send them to New Jersey or send them around to the Houston area, right? Can we route to vessels better? Those voyages take weeks to go from ne side of the ocean to the other. And our models are able to predict within 10 hours of accuracy when a vessel will arrive where it's going. And so we can predict not just where you should obviously send your vessels but we predict that by knowing where your competitors are going to be sending their vessels, right? And so that is something we're already doing for our commercial logistics customers there. They're not big enough at this point to change the totality of the problem we're facing. But honestly, this is a great driver to get us into the market and to make some of the bigger players become customers. So we're definitely trying to make hay of that and it's the right time to be going public. Absolutely

All right. Well, great. We hope you can help us. Now we're running out of time here. This has been a very dense hour, it's flown by. Unless we've got any other -- unless we've missed anything, if one of you like to throw something else in at the last minute here. I want to remind everyone about the vote. But I think we've covered a lot of ground. And if anyone wants to reach out, I promise that the BigBear folks and Luca are very approachable. So shoot us an e-mail and we'll pass it along to them, and hopefully, they can help you out. But before we go, I just want everyone to remember that the vote is coming up on December 3. You do not need to wait until then. You can vote right now to your shareholder of record. You probably could just go to your broker's website. It will take a couple of minutes. If not, there's some instructions here, which we'll republish again. And I believe once again that a couple of folks asked about the investor presentation. We'll include that link on the replay, which you can find either under the GIG ticker on Yahoo Finance or Bloomberg or on our website, and the replay of course, will be included in that article. But Raluca, Reggie, Josh and Brian, Jarrett, thank you, everyone, for making today's event happen and for everyone to watch. So have a great day.

Thank you. Appreciate it.