In this week's episode, I interview Dr. Brendan Klare, Co-founder, President, & Chief Scientist of ROC.
Brendan shares his personal origin story as an infantryman in the 75th Ranger Regiment of the U.S. Army and how his academic journey led him to co-found the only American-made multimodal biometrics and computer vision provider. Brendan explains why ROC is trusted by the U.S. military, law enforcement, and leading FinTech & digital identity brands to provide cutting-edge biometric technology.
Brendan explains ROC’s latest innovations and how they are unlocking new biometric use cases with speed and accuracy. We also discuss the rapid acceleration of deep fakes, generative media, the democratization of models and algorithms.
RESOURCES:
Connecting with Brendan Klare
Brendan Klare’s LinkedIn: https://www.linkedin.com/in/brendanklare/
ROC’s website: https://roc.ai/
Companies & Resources Discussed
ROC is the only American-made multimodal biometrics and computer vision provider. It is trusted by the US military, law enforcement, and leading FinTech brands. The company was previously known as Rank One Computing and rebranded to ROC in Q1 2024.
ROC Intern Program announcement
Scott Swan is the CEO of ROC. He assumed the role from Brendan Klare in January, 2020.
Professor Anil Jain is a University Distinguished Professor and Douglas E. Zongker endowed professor at Michigan State University where he has taught and conducted research for 50 years in the field of pattern recognition and biometrics.
Professor Sudeep Sarkar is a Distinguished University Professor, Chair of Computer Science and Engineering at the University of South Florida, Tampa, and Co-Director of the USF Institute for Artificial Intelligence + X.
Noblis is an independent, nonprofit organization with a proud tradition of serving federal clients objectively and with the highest caliber of scientific and technical excellence to help its clients make mission-critical decisions.
DARPA is the Defense Advanced Research Projects Agency and, for sixty years, has held to a singular and enduring mission: to make pivotal investments in breakthrough technologies for national security.
IARPA is the Intelligence Advanced Research Projects Activity. It operates under the Office of the Director of National Intelligence. It invests in high-risk, high-payoff research programs to tackle some of the most difficult challenges of the agencies and disciplines in the Intelligence Community.
FULL EPISODE TRANSCRIPT
Steve: Welcome to the PEAK IDV EXECUTIVE SERIES video podcast, where I speak with executives, leaders, founders, and change makers in the digital identity space. I'm your host, Steve Craig, Founder and Chief Enablement Officer of PEAK IDV. For our audience, this is a video first series. So if you're enjoying the audio version, please check out the full video recording on executiveseries.peakidv.com where you can watch the full episode. You can read the transcript and access any of the resources or links from today's conversation.
This week, I'm super excited to speak with Dr. Brendan Klare, Chief Scientist, President, and Co-founder of ROC. ROC is the only American made multimodal biometrics provider. ROC is trusted by the US military, law enforcement, and leading fintech and digital identity brands. Prior to co-founding ROC, Brendan received his PhD in computer science from Michigan State University. And prior to his academic studies, he served in the 75th Ranger Regiment of the US Army. I first met Brendan many years ago when I was leading product for a document-centric identity provider. We were adding automated face biometrics to our stack for the first time and one of the science leads from a recent acquisition told us we absolutely had to verify-- or we had to evaluate ROC then known as Rank One Computing. Welcome Brendan and thank you for making the time to be on the podcast.
Brendan: Hey, thanks so much, Steve. You're one of my favorite people in the industry. So I really appreciate this opportunity.
Steve: Oh, thank you, I appreciate that. Well, let's get started for those that have not heard of ROC. Can you tell more, like what's your typical elevator pitch?
Brendan: Yeah. Anyone with biometrics, identity, or video analytics requirements, we're here in the market as a trustworthy provider who can deliver to your team, top tier accuracy algorithmic solutions, best-in-class hardware efficiency, and second to none partnership support. We don't overthink how to do business. We are engineers at our core and we are really, setting ourselves up to be a long term partner for any of the challenges that you may face.
Steve: Excellent. Excellent. People may know your company as Rank One Computing, but you recently rebranded to ROC. Can you share what drove that rebranding and this new brand identity?
Brendan: Yeah, it's probably no secret to anybody who's followed us over the years that, you know, marketing was maybe not our strong suit in many ways. And that has changed here recently with some really incredible talent that's joined the team. You know, the rebranding, it's both to ROC-- ROC, it's a little simpler to recognize.
Some people don't know, kind of, the dorky background of Rank One Computing. So ROC, if you have a background in machine learning and pattern recognition, there's a receiver operator characteristic, which is a way to measure and visualize trade offs of type one and type two errors.
So that was the genesis of ROC, Rank One, if you're in an image retrieval scenario is, you know, the top search result that you get as well. Some people don't, you know, really resonate with that all the time. If you're not sort of a dork, you know, you just spend a lot of time in those circles. And if you think about, you know, where we are, we're in Denver, Colorado, you know, we have company offices also in Morgantown, West Virginia. You know, the mountain is, you know, a real pillar of strength is a lot of long term presence that are there. A lot of reliability that is all part of the genesis of our name right now. And so we're really, you know, doubling down on that, trying to make it a little simpler to recognize us, just be a little cleaner and better looking. We have some pretty incredible people who've joined us recently to try to make our market presence the same as our technology, you know, which is very clean and sleek.
Steve: I got to say, I love the new branding and I'm a bit jealous because I called my company PEAK IDV and I've got mountain imagery in there as well, but you've got that up arrow, very simplistic designs is very, very cool. Well-- very well done and congrats on the new brand identity.
But before we go deep into ROC, I'd love to dive a little bit deeper into your personal background and your origin story. Can you share about when you decided that you were going to join the Army?
Brendan: I was fascinated with math and science and that's something that, you know, it's been a big passion of mine. And, you know, I think if you look at a lot of younger people out there who are, sort of, you know, transitioning from teenage years into adulthood, you know, there's-- it's hard-- it's hard to take everything seriously and find everything interesting. And, you know, there was a series of events happening, learned about the Army Rangers, the challenges that are there. And, you know, I don't think that I was ready to take that-- full advantage of the educational resources, you know, that are here in this country, which are phenomenal. Needed to get some stuff out of my system, jump out of some airplanes, blow some things up, kind of, gain an appreciation for the value of an education as well.
So that's how I ended up in that journey. I'm pretty passionate and intense person for those who know me. So, you know, I think there's a lot of 18 year olds, 19 year olds, 20 year olds out there that probably could really benefit from it.
Our nation's armed services are unbelievable, absolutely incredible. All the intangibles you get from that, you know, I think it's hard in some ways to put a pause on your life, to put a pause on your career, put a pause on your family, particularly with the events that were happening around that time. But even just from a selfish perspective, if you don't do it in the service of others, which to me, I come from multiple generations of veteran service to our nation's armed forces.
I think it's four consecutive-- I'm the fourth generation of combat experience, you know, on behalf of our country. But even outside of that, just for personal growth and structure it's a really valuable thing. And that-- that's how my experience played out, fortunately as well. And I know that's not the case with everybody. I mean, there's a lot of different experiences that can happen. And, there's some people, you know, they never come back or there's some people that are never physically the same or mentally the same. So I do feel very fortunate that that was not the case with me, you know, it's a real positive experience. So I often refer to it as the most fun I never want to have again.
Steve: We appreciate your service, Brendan, and for your-- your family-- family's heritage and the armed forces. And when I looked at the time that you were a Ranger, 2001 to 2005. I mean, that was a really transformative year for us. We had the September 11 attacks and the Patriot Act, the war on terror operation on Iraqi freedom, the war in Afghanistan.
It's like 9/11 was one of those watershed moments for us. And it had been almost a decade that we had had any major conflict. You know, the Soviet Union had collapsed. The Gulf War had settled. Do you remember where you were when we had those attacks on our soil, on 9/11?
Brendan: I do. I, you know, I guess most people do. I had a bag packed in my Army recruiter's office, about to ship to basic training at, then Fort Benning, and actually I had a medical injury, sort of, unrelated to anything and I was determined to ship and they were telling me, “well, hey, I think we have to tear your contract up. I don't think that you're going to be able to ship, like you're going to redo this whole process again.” And, as we're talking about that, somebody comes running in, you know, one of the recruiters is like, “turn on the TV, turn on the TV.” And, you know, that was 9/11 for me in that moment. And it did turn out that my contract was null and void. So I both joined the Army before and after 9/11. I had to go through the whole process again.
Steve: Yeah. I was actually working at a bank. I was a bank teller and no one came into the bank that day; it was very quiet. And then everyone was just glued to their TV, at least from-- I was in California, but I feel like September 11 has really changed the global geopolitical landscape, it's changed identity verification, how we think about trust in the world. I understand you also did three deployments in Afghanistan. How did your time in Afghanistan influence your career and your future interests?
Brendan: Yeah, you know, directly, probably very little. So, you know, I was an Army Ranger and the truest sense of the word, you know, I was an enlisted infantry man, 3rd Ranger Battalion. You know, we-- with the front lines of everything that happens there, my experiences are probably lighter than a lot of my colleagues. You know, I have some guys I served with ended up on different paths and different missions that saw things that I never personally saw. Indirectly, the impact of that was tremendous. The adversity, obviously that you face in those scenarios, is high. The amount of professionalism, particularly when you're working with the special operations community, you know, within the US Army and the Department of Defense, it's almost impossible to describe for anybody who has never worked in those environments.
And we're-- Army Rangers are a tier two element. So we do our own direct action missions. We also support tier one elements, you know, like Delta Force and Seal Team Six and working alongside those guys, you know, who are career professionals in that community, you realize how serious things are. And if you want to be good at something, you have to be very dedicated to it, and you can't just casually take interest in something. So those are the things that stuck with me. And I think if you look at running a technology business that bit is the same. You know, if you look at how the structures are set up though, and like special operations community, or just, you know, a larger infantry, you know, division, there's a complete and utter deference to leadership. There has to be, you do not have the liberty when, you know, the proverbial blank hits the fan, to be questioning who's in charge, you know, to be having disagreements of opinion. And so that's a world where it's just like you fall in line and you do exactly what you're told. Or if you're a leader, you tell people what to do and everybody operates in complete concert with one another.
In a technology company is a little more chaotic, I would say by design. And that's really our culture here is that, you know, at the end of the day, there's certain people that are paid to make certain decisions. And when we get to that point, we all fall in line and do it.
But what leads up to that is intentionally a little chaotic in a good, polite, professional way that-- we have a motto here at ROC, for example, ‘hang your ego at the door.’ So we go into a meeting and you could be the CEO or chief scientist of the company, or you could be somebody who's been here, you know, for a year and, you know, got out of college recently. The right answer is the right answer-- period-- point blank. And you're expected to advocate for the right answer. Sometimes it's subjective what is or is not the right answer. So those are also things that I would say, you know, sort of the negative example, you know, I learned through the military is that that is not the right way to run a lot of organizations, even in an incredible unit like the 75th Ranger Regiment. There's some of the best people I've ever met, and there's some people that are really difficult to work with, but they're also incredible at certain aspects of the job, and, you know, you can see areas of bad leadership with some of those folks, and it really-- it paints a picture of how you should never lead, particularly when you're not into a military environment.
And we see people, even who don't have military experience, who try to run companies in an authoritarian manner, which again, at the end of the day, somebody has to make a decision. People need to get behind them when that time comes, but you know, they think they know everything, they want everybody to just do what they’re told. There's some aspect of ego that gets involved in it, and it's really harmful to so many people. So, I'd say indirectly to that-- that was pretty impactful to get such an exposure.
Steve: Leadership is a common theme. When I speak with those who've served in the military and they've gone into the commercial and private sector, that those are some of the best lessons they learned. But, to your point, you can’t always take that hierarchical authoritarian approach as you go into a technology company.
Fortunately for you, you transitioned from the military life into academia, which is a completely different area. You start knocking down degrees. BS and CS., MS and CS, and then you went on to be a Spartan PhD candidate and eventually earning your PhD with Professor Anil Jain. When did you first decide, “hey, I'm just going to go on a tear and get all the degrees?” And then when did you decide, “hey, I want to dedicate my life to biometrics?”
Brendan: Yeah, great questions. Yeah, when I got out of the Army, I mean, I was 100% focused on getting my bachelor's degree in computer science. That's where I had started before I joined the military. I got in trouble in high school because the computer science teacher there, -- which I was fortunate that we had that, you know, type of education in high school, but -- they didn't know too much about it. So it's like within a few days, it's like, you know, more than the teacher.
Not that I'm anything special in that regard, but, you know, I both loved it and got in trouble with it. But having that initial exposure, I knew exactly what I wanted to do my professional career, no doubt about it; I wanted to develop software. So getting out of the ranger regiment, getting into education again, I went to university of South Florida, which is a great place down in Tampa. Did 90 credits in 24 months for my bachelor's degree to get that. Thought I was going to stop there and just go right into industry. My senior year, my bonus taken classes, equivalent of senior classes, Professor Sudeep Sarkar at, University of South Florida, who's one of the smartest people anyone will ever meet, he's still very active down there. I took his computational geometry class, which it was really, sort of, a conduit into computer vision technologies. He actually-- he saw my interest in it and I wanted to develop software to track pitches coming from a pitcher in baseball, because-- this is before they, kind of, had that ubiquitously, you know, and all the sports broadcasts and stuff.
And he set me up with all the camera equipment and computers and we did a pretty cool job with it. I wouldn't say that that would deploy, you know, to the Colorado Rockies, perhaps-- but anyways, that was my conduit. So I decided to do a master's degree and that's where it just gets so fascinating.
So you start taking those courses of machine learning. You start taking those courses-- courses in computer vision, even just, you know, the advanced, you know, courses and operating systems and algorithms. I was so fortunate to link up with professor Anil Jain. If anybody doesn't know who professor Anil Jain is-- he has a 50-year history now of publishing pattern recognition and biometrics research, some of the first ever work on automated fingerprint algorithms.
He's one of the nicest, kindest, most naturally passionate people you'll ever meet. He really treats all of his students like they're children of his in a good way, though, sometimes-- some need some-- the true parenting on it. But you know, it was really a partnership it felt like in the PhD study. We achieved so much in the world of face recognition together.
I learned so much. And then even when finishing, you know, he, I think, is a little bit unlike some other professors-- and this is something all professors and I think all professionals period should take note of that. What happens to his students after they graduate is equally as important as what happens when they're with him doing their studies.
And there's some professors, and I don't mean to speak negatively in this regard, but I mean, it's a stressful job. It's a focused job and, you know, publication count, you know, your next grants and research funding is all that matters. And when the student's done, I mean, they're kind of on their own from there.
And that's not the case with Professor Jain, and I'll talk about that here in a minute. So we'll talk about ROC CEO Scott Swan a little bit. And, even there, you know, there's really a small world that I've been fortunate enough to live in that started with working with Professor Anil Jain. So everything we're going to talk about probably from here forward about ROC and who we are right now, it has professor Jain's fingerprints on it and I'll forever be grateful for him.
Steve: That's great. That's great background. And that leads to the next question I have is around the founding of Rank One Computing. And I saw a post you did on LinkedIn a couple months ago, where you're talking about the ROC intern program. And you made reference to your time interning at Noblis. Can you share how the company started?
Brendan: I was always going to start a company. I mean, that much was clear. I was one of three co-founders of the business. Some of the exposures you have-- so you sit in academia and you see the smartest people on the planet solving the hardest problems on the planet. And you realize that they are changing the world. But they can only take it 90% of the way, typically, both because it stops getting interesting after that, and it's a lot of work to do the minutiae.
So like to build a face recognition algorithm that achieves very high accuracy marks versus building the same face recognition algorithm that now can deploy as usable scalable software to a host of use cases and customer personas are two very, very different things.
So, that was one bit of exposure that made it clear that, to really solve some of societal's problems and some of the opportunities that are out there, an academic career is not going to be able to do that.
I went to work for a company called Noblis, who's a, really impressive-- I'd say probably it's like a think tank, they're a nonprofit in the DC Metro area. Through doing that, got a lot of exposure to how these agencies use and procure various biometric and computer vision technologies. Probably the biggest thing that I saw that was just-- okay, there's absolutely an opportunity here is how overpriced and poor customer experience there was in these industries.
You have agencies that are spending tens of millions of dollars on technology that can't even talk to the people that are developing the algorithms. They can't even get access to new algorithms. An ‘aha’ moment was when one of the most important national security agencies out there is using 10 year old face recognition algorithms, despite paying maintenance fees every year. So they're paying for updates on this, but they're not getting algorithm updates. They're getting these frivolous software updates. And that just felt unacceptable. It felt unethical. And just from an engineer's perspective, there seemed like a smarter way to build and deliver technology. And so all of those were the factors that came together.
Steve: That's great. I didn't have all that background, Brendan. I know you started the company in DC and then you later moved it to Colorado, to Denver. I think the first time you and I met in person was in 2017. And I actually looked on my phone, I took pictures of-- we went to Work & Class-- we went to this place called Whiskey Foxtrot, something-- something (Whiskey Tango Foxtrot). But anyways, I didn't take any pictures of us, it was just like the beer and the food, which shows my priorities for the time. But back in that time, there were seven people total for ROC. Like, everyone fit into a conference room. What does the team look like now in 2024?
Brendan: Yeah, great question. And, yeah, I guess as a bit of an aside, man, Steve, I really did always have so much fun engaging. We didn't know each other before our paths collided then. And, for anybody who listens to your podcast that hasn't gotten a chance to work with you, it's just-- there's very few people that both are really smart on product and technology, but are just also easy to do business with. And it's-- it just makes life better and more productive. So, it was cool that you believed in us at that time. And we did have fun. I actually also remember going out to San Diego and I wish I remember the brewery we went to, but you know, between Denver and San Diego, I mean, you got some of the best beer-- I'll say in the world. You know, maybe some say, “ah, just the US,” but anyways, we share similar priorities, my friend. Good memory on Work & Class. If anybody comes to Denver, that's still a great spot.
Yeah, we were seven people back then. You know, our first five employees were all engineers, all science background. And was actually that way for about, I think, a year and a half. Next two hires there, when we were the seven, it was-- now our chief financial officer as one of them. And who is now, our chief partnerships and privacy officer who's our general counsel is the other. So a CPA and a juris doctor, a lawyer.
So it's always been very skill oriented here at ROC. We're over 50 employees right now, which, you know, I know in some companies where there's, you know, the kind of the big tech booms might not sound big-- we've never taken investment. We have bootstrapped this entire thing from the ground up. We've been profitable almost every year that we've been in business.
So all of this is measured growth-- if you look at our team right now-- whereas we had to do a lot more sort of jack of all trades aspects of it. Go look at some of our old marketing, to get back to where we were at the start of the podcast, to today and you get real experts in those fields, it's really incredible and powerful to see. And we'll kind of talk about my transition into the chief scientist role in a bit, but I'm getting people who are really good at stuff, able to focus more squarely on their strengths and not generalize as much. It's very powerful.
There's a lot more, sort of, over professionalism to the company and we've always been professional acting, but if you, kind of, looked at us, people might not have, sort of, felt that in some ways, though. The-- obviously the-- the tech worked extremely well from the beginning and part of this growth to-- yeah, we have three people, maybe four who've left our company in the almost 10 years that we've been in business. The retention rate within our team is sky high and that's no, sort of, random occurrence. It's every bit about, sort of, the DNA of who we are and the long term mindset that these are 10, 20 year problems that we're focused on. And so we-- we really-- whether it's the fact that we're employee owned, so everybody, you know, has pretty good stake in the business. Where, you know, talking about the hanging ego at the door, trusting each other, all those aspects are the same. So while we've grown a lot since then, just about everybody you knew then, is still here now and we're just a larger, more capable organization able to deliver a lot more to the marketplace.
Steve: That's a testament of the culture you've built. And what I really admired during those conversations, when we were getting to know each other and our companies were doing business was you had this, this mission-first, people-always approach, which no doubt came from the military, but you were also adamant that you didn't want to raise venture capital.
You're not looking for some exit. You didn't want to just blow this up and quickly go to market like some of the companies that have taken on lots of venture. Your ethos was to solve tough problems, and you just wanted to work with people that you wanted to work with. So I really-- really admired that.
How have you been able to maintain your values and culture? You alluded to being employee owned, are there any frameworks or approaches you've taken to be able to maintain that culture?
Brendan: Yeah, thanks for the compliment and the observations there. I mean, it is something that we take pride in, you know-- we take seriously-- we hope to continue to keep this all together.
You know, when you grow as much as we have grown and, you know, just in the last three years, you know, we're about three times larger and employee count than before. And, you know, frankly, it's hard to grow more than that and keep this in place. So as much as we've grown a lot, we've tempered-- you know-- sort of-- temptations to grow a bit faster than we could have via the bootstrap approach that we're doing as a business model.
You know, the 50 people who work here now, so many of us have worked together prior to ROC, and that really helps. Or maybe we didn't directly work together, but we worked in sort of partner organizations. So there's a lot of known trust, you know, between ourselves within the company.
So that helps, you know, when you have an engineering background, you know, it could be good and bad in terms of like the business model that's there, but it does become very problem oriented.
And, you know, the things that I believe can often really derail a company. There's a couple of things being manic behavior. So, you know, you hire, you know, “Hey, we just hired a hundred people.” And then, you know, two years later, like, “ah, we just fired 30% of the workforce,” the psychological toll that that takes on people is tremendous. And that's something that, you know, you never know what's down the road, but you know, you really want to avoid that, you know, also with customers, I mean, customer churn is a really difficult thing.
If you look at our industry, you know, even, you know, with your, former employer, where we worked together, who is still a thriving customer of ours, and we're just so thrilled to be able to support them-- the lead up to, you know, really where you all started taking technology operational and then from there to where that operational technology is generating meaningful revenues, it's a very long period of time. And so to be able to take that long term approach is important. And to be able to recognize that the customers are going to take a long time to build everything and to support them along the way to be a customer service organization, that if you are working with us and you have an issue, you're going to talk to the person that built the technology-- period-- point blank.
And there might be a business person involved in that conversation, but they're a facilitator, and that's what they're there for. They're not going to have the answer for you and they're not going to pretend to have the answer to you. You're going to need to get that. So, as we get bigger, I'll talk about some other, you know, reasons why we're able to maintain this, but we do see this as scalable. We see this as common sense and we don't think it's that hard to do, personally.
Steve: Yeah, to summarize what I just heard is you really are making sure that the pyramid is employees, customers, shareholders. A lot of companies in corporate America today are shareholders, sometimes customers, and employees last. I think if you take care of your employees and your culture, it'll take care of the innovation and your customers, and then ultimately the shareholders will benefit.
I want to transition a little bit into something you've alluded to a couple of times, which is you have a CEO, Scott Swan. You were the CEO from the company's formation. And then at about the end of 2020, you became the chief scientist. Can you tell me more about that transition, like how you met Scott and how the company might have changed under his leadership?
Brendan: Yeah. I mean, one of the best things I ever did was fire myself as the CEO. It was both for the company and myself. You know, the phase that I was the CEO, I had to be the CEO. That's how we got to where we were at. We had a lot of people focused on the technology, that's my background as well-- I'm a student of business, though I've never-- you know, I don't have any degrees in that. But you know, we needed, a real executive leader. We needed somebody that understood the market demands out there beyond just face recognition.
We needed somebody that understood how different organizations purchase. And all of these things is, Scott Swan. I've known Scott for almost 15 years now. He was at the FBI. He spent almost 20 years at the FBI. He rose from being a copy clerk to a special assistant to one of the executive directors of the FBI, which if you don't know that organization, that's a powerful position that he was in.
I remember actually when I was an intern at Noblis and Scott comes-- and Noblis did a lot of work with the FBI-- and I heard like, “Hey, this guy's coming, you know, from the FBI, everybody, you know, sort of tighten your ties,” you know, this is back when people wore ties there. You know, here comes Scott and he looks fairly young and he still looks younger than he is. He-- he's a little older than people probably might think. And, there's a whole, you know, trail of people behind him. And it's just like, I don't know who this guy is, but, you know, people seem to really want to move heaven and earth to support him. And, I eventually got to work with him directly after the Boston Marathon bombing, which is, you know, another tragic incident.
You know, when that happened, there was, you know, over 100,000 images that people had taken around the finish line and I think 10,000 videos, because obviously everybody's filming the marathon. And, you know, after it happened and we had no idea who did it, you know, the FBI is, you know, reach out to the public and be like, please send us anything and everything that you have. And the amount of imagery that they had received, it broke all the systems. This again, when I talk about hardware efficiency and scalability, this was an ‘aha’ moment for us as well. But I got to work directly with Scott on a major issue study where after the fact, after that whole, you know, incident cleared, you know, the FBI basically said, “how can we prevent this from happening again?”
And so, Scott led a task force that Noblis was part of to go and study all the market technologies out there for computer vision, machine learning, really video analytics. Fast forward to today, a lot of the knowledge that was learned there-- a lot of the requirements-- they haven't changed what people wanted back then-- they still want today. And some of it hasn't fully been solved and delivered yet. That's just like a sliver of my experience working with Scott. And he lived that world for about 20 years. He spent almost five years at our direct competitor after that in a very senior leadership position. I had talked to Scott every bit along the way, hoping that our worlds could collide. And, you know, we were kind of a small fish. He was a big fish, but he always had an interest in the stars aligned and he came here and you see it.
Steve: Do you recall the moment when Scott said yes, you knew that he was going to be joining?
Brendan: Yeah, it was like a series of moments, but yeah, you know, when something's going to happen eventually. And it was just like, you know, probably the biggest feeling was like, I felt confident in it, but it's just, you know, I hope that we can deliver for Scott because, you know, he was sort of leaving a lot behind to come to us and he saw that vision. And two, for him to kind of realize like, “Hey, we're not going to take investment for you to achieve your vision.” We want to maintain control of this company so we can be in this for a long time to come. So yeah, it was really cool. And I think if you were to ask him today, he'll say it's one of the best professional decisions he ever made, but I'll let you ask him maybe at another point.
Steve: Yeah, that'd be great. I just-- in my mind-- I thought you going “game on, here we are, ready to go. It's time to grow the business.”
Brendan: Oh, yeah.
Steve: Well, let's talk a little bit about your products today. What are the core modalities that you focus on now?
Brendan: Yeah, I mean, we're still obviously a face recognition company, top tier in the NIST rankings. With that, we're still-- all of our algorithms-- we are top in accuracy, efficiency combined. Nobody can build algorithms that are as fast and scalable as ours. That's just born out in the benchmarks. Space recognition is still a big area of focus. Liveness within it, presentation attack detection. So we're ISO 3017-3 level two compliant. I look at my notes to make sure I say that correctly. Facial analytics as well. The ISO 29794-5 face image quality standards, which are now increasingly required for a lot of passenger travel. And even if you look at the identity industry, you know, for doing identity proofing, you really need to control some of the capture process. So we have very strong algorithms there. We're in the NIST SITC benchmarks on that. Fingerprint recognition-- I mean, that's been one of the most exciting developments. A year and a half ago, we had no fingerprint algorithm available to license. Just a year and a half later, we're top three in the world in accuracy per the NIST benchmarks.
We just published on the NIST latent fingerprint recognition benchmark. When we were published, our search time for a 32 million image database was 500 times faster than any other vendor in that benchmark. And we had top three average error rates in that. That's 15 seconds to search a 32 million database versus 42 minutes for the next provider.
So that's the next generation capability if there ever is one. Iris and tattoo recognition are capabilities as well. And video analytics-- I mean, this is when Scott came in day one, he said, we're going to be a video analytics company. We're not just going to be a biometric provider. We are going to build a computer vision algorithm.
Gun detection-- weapon detection is one of our biggest areas of improvement recently. Obviously, there's a lot of need for that unfortunately in society-- we have some very important customers in that space. I think people are going to be really wowed by the capabilities the next year, in particular, as it's improving license plate recognition, vehicle make. model. Quietly, we are one of the strongest capabilities in industry in that area.
Person recognition, person re-identification. We do a lot of special research projects as well. We're very selective about this, but whether it's DARPA, who has been a client of ours since day one in business-- well, let's say like day, you know, 60 in business-- IARPA and a couple of these three letter agencies, we do a lot of specialized research projects for them.
Steve: Wow, your product stack has exploded with functionality. You know, the fingerprinting technology, can you-- can we-- double click on that a little bit? Like, what are some of the new use cases that that opens up for you guys?
Brendan: Yeah, I mean, if you think about-- I mean, you know, it's the standard use cases that exist, but now it's opening up all new products possibilities as an industry, as a whole. You know, even just-- I mean, you have these unsolved latent databases where people have committed heinous crimes. You have a capture of their fingerprint-- it's in a database-- the technology is so slow that you're really limited it to the times that you can search those databases. So, being 500 times faster, I mean, you can double, triple the size of the database and it's a drop in the bucket; it doesn't matter. People that are routinely fingerprinted for background checks in theory could all be submitted against latent unsolved case files. We're the lowest error rate on the IARPA nail to nail data sets, which is a suite of about 13 different sensors for contact and contactless fingerprint capture.
So I think we're seeing some of this move into the ID sector on the private side for fingerprint through contactless capture on a mobile phone. So that's an area that we're very strong at as well. It also really helps because, you know, our big goal, we really think that we're the most logical solution for the large government, national security agencies, critical infrastructures and being a multi modal biometric provider really allows us to go and service those markets.
Steve: In the origin story, you were describing the market back when you got going was just a couple of large foreign companies. You know, these are very sensitive national security projects in some cases, and one of the prevalent things in your marketing is the notion that ROC-- It's just an American made technology. How do you see companies latching onto that messaging? And what is-- has that driven a tailwind for Rank One-- ROC? Can you describe how that's worked out for you? The messaging and…
Brendan: Yeah, absolutely. And, yeah, and to be clear, I mean, we're global citizens. So, we love and appreciate everyone who's out there, you know, at the same time in the US is a massive economy. The US government is such a critical piece of how the whole world itself functions. So it's natural to have a desire in government agencies for US tech. You know, as the AI boom is happening right now, I think there's several things occurring that people really need to take note of.
One is, we are not spending the most money on artificial intelligence and computer vision, the United States, even Europe. And, at least, you know, if you kind of normalize for like what the cost of a given employee is, you take United States and Europe and add them together and it's going to be less than China.
They're so aggressive on this front. And they've also made it clear that they don't agree with the current world order. And we're seeing this, unfortunately, geopolitically right now as well, that there's a challenge to Western civilization at the moment. When you need to trust your technology, it really is important that you know who is developing it.
So just being US owned and quote-unquote developed doesn't quite do it. I know companies that are US companies that develop stuff overseas. So for us, it's, you know, all our employees that are building the technology, it's all people that sit here, you know, in the United States, that build it. There's a growing threat of poison AI that is out there. And if people who haven't heard about that, you know, these models are very, very complex and complex systems are hard to validate. And if they're hard to validate, you can put back doors in them. And this is not like your traditional software scanning, where you can sort of search for those vulnerabilities.
People can set it up such that somebody-- let's just say that you have a border security biometric system, and maybe you have a foreign piece of technology that's in there. And maybe certain people are actually set up such that they can clear right through and not match against certain watch lists that are out there. Or, they can assume multiple identities at once, you know, while crossing through borders.
That's but one example of things that can be done. How do you mitigate against that? It's really hard. He can't actually test against 7 billion people, obviously, you know, on these algorithms. So, really, one of the first and foremost things you can do is work with somebody that you trust. A company that has first-principles, a company that has a track record of not doing these things.
You know, you can search a lawsuit center out there against some of the biometric providers. They're very interesting. The ones that exist, you know, there's accusations of Russian software developers putting back doors into some of the most important biometric systems that are out there. There's, you know, other cases where just kind of poor ethical decisions have been made.
So for us, when we-- we talk about being US owned and developed, I mean, obviously, here at home, this is the number one market that we want to be strong in. So that's a key part of it. And then another part is just, you know, really continuing to paint the picture of we're a very simple company to understand. There's not a complex structure here. And so with that, we think it affords higher trust.
Steve: One of the points you just made around technology, is the ethical use of technology and not building in these back doors-- really being transparent in what you're creating. I want to call out something that you have on your website, which I haven't seen many companies do, which is publish a code of ethics.
With the rise of AI and there's all this talk about government regulation. I think there is an important part of that-- that's just self regulation and self governance. What's your company's commitment? Maybe we can summarize your high level code of ethics for the audience.
Brendan: Yeah, I mean, at the end of the day, our technology is meant to do good and to not be harmful. We think that, you know, in absence of a lot of sort of legal and regulatory, sort of laws out there, that there are a lot of informal standards that can be adopted. There's a lot of misinformation on the topic of the technology. So, we paint a set of first-principles and that is meant to be, you know, general sandbox to play in that-- you don't have to rethink every scenario that exists-- that we can point back to it. We believe we were the first company to publish this.
I don't think that's something to necessarily be proud of, you know, this is something that maybe to be a little more ashamed of as an industry that this is not something that is thought of more though, there's no shortage of virtue signaling that are out there on any given topic. You know, people want to take the high road and kind of follow the sheep on like, “Oh yeah, this is what you're supposed to do in this scenario, of course.” And, hey, are we going to make a bad decision? Sure, we're susceptible to that. We really try not to. So when you have a set of first-principles that the team really buys into starting at the executive level. But then from there, every scenario on a case by case basis, you can further scrutinize it. You also empower, you know, decision making within the executive team. There's not convoluted decision structures. So we have open, free flowing information to talk about this. We think that all helps to limit the misuse of the technology, to know who our customers are, to know what they're doing with the technology and to trust professionals that have thought through these things. You know, a lot of it, too, is a starting point. You know, if you look at our software licensing agreement, we say it-- period-- point blank, “we can revoke access to our software if you violate the terms of our code of ethics.” So that then, in and of itself, if you're a business person entering a business contract, you do not like that. That is uncertainty, right? That's ambiguous. These are not, you know, sort of codified laws. This is our set of ethics that we present. It starts a conversation though. So now we can talk to you. “Why are you concerned about it?” You might have a very real, simple, innocent, you know, reason that that's just, you know, bad for business to have that kind of uncertainty that allows us to understand your use case even further. Ninety-five percent of the companies out there-- doing really, really wonderful things. Or the government organizations, you know, it's kind of those 5% that you got to watch out for and be very careful.
Steve: On the ethical use of technology, one of the hot topics right now in the digital identity space is the use of deep fakes. And there's AI for good, there's AI for bad. We're all very concerned about AI for bad. This generative media democratization that’s been happening is really starting to affect a fraudster or criminals ability to trick these systems. How do you see ROC fitting into this whole Gen AI ecosystem? You alluded to liveness detection, but are there other ways that you're thinking about the problem?
Brendan: Yeah, it's a great question. I mean, obviously, deep fakes has really captured a lot of interest. For those who don't know specifically what a deep fake is, you know, it's kind of creating a synthetic identity or taking an existing identity and, you know, sort of recreating it and being able to do what you want with it.
We get asked this question, almost daily. You know, like “what do we do about deep fakes?” You know, it depends on who you are, but let's say for the bulk of the customers and sort of in the banking sector, know your customer, ID proofing-- I think, you know, Steve, a lot of the market that, you know, you're most familiar and most catered towards.
And there's a lot of fear, you know, of, “are we just going to get blown away by, you know, the advancements here?” If you look at-- so say somebody creates a deep fake, there's really two ways that they could use it with an identity system. One is a presentation attack. So yeah, I put it on a tablet or a printed photo and I hold it up to the biometric sensor, you know, like a mobile app-- do an ID proofing-- and use that hopefully to let that fake identity be accepted.
The other one is to do an injection attack. So to sort of bypass the cameras, put it into the feed. So it's not a replay or presentation attack. You know, you're kind of directly inserting it into the camera. In the first example, like presentation attack, you know, you said it-- liveness. I mean, that, that is the solution to it. We're seeing that liveness is increasingly looking like a solvable problem. And there'll always be, you know, a little bit of a cat and mouse game there because of how sensors change and how, sort of, fraudsters, you know, look to present, sort of, replay attacks in different ways. But, you know, overall it looks like the accuracy of these systems from independent benchmarks are quite good. And, you know, those vectors are being eliminated.
Injection attacks, I mean, this is an area that we've really been focusing on lately. As a full stack software provider, you know, you need to be able to trust the source of the camera that exists. There's various methods-- various tools that are out there. I'd say that this is both something that companies, like ROC, need to take leadership positions in and deploying technologies that prevent people from being able to bypass cameras and insert virtual camera in instead. Also, I mean, the kind of the two main players, Apple and Google, I mean, they, you know, control the mobile market. They are critical to the problem as well. And, at some point, you know, need to be able to provide a higher level of assurance that you cannot bypass their cameras without it being known at the application level of a given system.
Steve: It's fascinating. I'm curious, with your history that we just discussed -- you've been a leader in detection, video analytics, computer vision -- do you see a path for ROC into the creation of media? Because like it's generative AI and creating content in your future?
Brendan: Yeah. I mean the short answer, like for generative AI and ROC is no. That's not something that we'll get into. You know, behind the scenes, under the hood, you know, there's a few areas that it can have relevance and may already have relevance, without getting into too much of how the sausage is made. But in general, you know, we're discriminative AI company. You know, if you look at machine learning literature, we develop discriminative machine learning methods, which, you know, really seek to infer information versus create it. So-- short answer is no-- I don't think that we'll be getting into that.
Steve: Brendan, we're almost at time, and if you've seen any of these episodes of executive series, I like to go a little bit beyond what you can see on a LinkedIn profile. I'm really curious what you personally like to do in your free time. If there's any organizations you like to support, can you share more about you?
Brendan: Sure. Yeah, I mean, I spend a lot of time with my family. I have an amazing wife, Jessica, and I have three amazing children. So that obviously is a big focus. Outside of that, you know, here in Colorado, skiing, mountain biking. Right now it is both skiing and mountain bike season. So that is quite nice.
You know, our big passions-- love live music-- into all those hippie jam bands. You know, Phish and Goose, Billy Strings, whatnot. Cooking food, lifting weights, just trying to live a healthy, happy life-- and balance, that's a big part. You know, you talk about why we're here today, this is stuff that for all of our team, we really, really focus on. We shut down for two weeks over winter break. We force people to take time off as a business. This is a long call and if you don't have balance in your personal life, it's just-- it's hard to sustain good creativity on the professional side. So, that's sort of an organizational thing as well.
Steve: Absolutely. Well, thank you for sharing. Well, as we wrap up on the conversation today, for those that are listening or watching, what kind of conversations would you like out of this for anyone to reach out to discuss anything that was shared?
Brendan: Yeah, I mean, talk to us about your problems. You know, there's a lot of use cases out there that are serviceable out of the box. And it might be the case for us that your problem is something we have an out of the box solution for. Might be the case that you have a more complex problem and we might say, you know, “thanks., but no, thanks.” But we'll give you a lot of information and building up to that feedback. Or it might be the case that you have a complex problem that we don't have an out of the box solution to. We have a lot of the components for it, and we actually agree that it's a problem that needs to be solved, that there's a real business case for that. And we'll dig deeper, but it all starts a conversation. And we really believe that we're an easy company to talk to, that everyone here is very approachable.
Probably the best way to get in contact with us-- just go on our website-- fill out a little form and somebody's going to get back to you in very short order. If you ever do start working with our technology, one of the first things you'll realize is we're there every step along the way. You're going to get to talk to the people who build it.
These are tough-- tough systems to build as you know, Steve, you've been on the front lines for a long time and it takes a lot of cooperation and partnership to do it. And, you know, we look forward to talking to anybody that wants to see if there's a way to work together in that regard.
Steve: Excellent. Well, I'll be sure to put your website URL and the resources as well as the contact form.
Brendan, thank you so much for taking the time to speak with me. Your journey with ROC is really inspirational and thank you for founding one of the foundational biometrics companies in our industry.
Brendan: Appreciate it, Steve. Congrats to you on all your success as well. And I'm just really-- really fortunate to get to connect with you again.
Share this post