Hello, everyone. Thank you for coming. We are almost at the finish line. But you know the old axiom, we always try to save the best for last. I have had the pleasure of knowing this man for many, many years. In fact, we share some history together. It was the only one-hour special I have ever done before his IPO. That was one of the great stories. We had no idea what you were going to go through, both on the upside as well as the down. In any aspect, I believe that our next presenter has some extraordinary disruptive technology that will definitely benefit everyone here in the U.S. I think that anyone who came into this conference saw what this man has built. With that, I would love everyone to give a warm round of applause to the President, Chairman, and CEO of Knightscope, Bill Santana Li.
Thank you, sir.
Of course.
All right, before Chris, you take off anywhere, round of applause for Chris.
Thank you.
For paying the ransom now that he held us hostage here for three days. All right. Welcome, everyone. My name is William Santana Li. I am the Founder, Chairman, and CEO of Knightscope. We are building the nation's first autonomous security force. This is a unique combination of hardware, software, and humans to see if we can make the United States of America the safest country in the world. That last statement sounds a little outrageous. The crazy thing is, having now worked on the problem for over a decade, we actually have the confidence and the plan to do it. We just need to execute. Okay? I am not going to PowerPoint you to death, so some interesting stories here to share with you. Normal safe harbor stuff. We are building this new managed service provider for physical security that has never existed.
We unveiled this, at GSX, which is the largest security conference in the world, to 16,000 security professionals a few weeks back last month. We got three reactions. First, absolute excitement. At one time, we had, I think, five chief security officers and two federal agencies in our booth all at the same time. Then we had speechless, do not know what to say, and then a lot of cranky people. We are going to talk about some of the crankiness. The first one is we unveiled the K7 autonomous security robot that you saw on your way in here, to the conference. This is a completely unmanned vehicle.
I don't know how many guys have come up to us the last few days, "How do I get in?" I was like, "You missed the whole point." This runs 24/7 to patrol very large perimeters up to 10 mph . It provides a physical deterrence. What does that mean? If I park the marked law enforcement vehicle in front of your home or your office, what happens to criminal behavior? It changes, right? Just physically being there with a 2,500-pound vehicle, strobe lights going, says "police" on the side or "security," it may greet you. That usually will stop half of the shenanigans. This is an important node of data that comes into the software orchestration platform I'm actually going to walk you through today. This generates about 90 terabytes of data a year that no human's going to be able to process.
The first handful, you saw MIN777. We don't have vehicle identification numbers. We have machine identification numbers. MIN777 is out in the front of the hotel. That's our first vehicle. We're setting up production now. We got our first handful of clients. We've got a very large wait list for this. We'll start scaling this up for 2027. We said this vehicle, we said we'd unveil it at GSX, then we're going to start shipping in fourth quarter of 2026, and that's exactly what we're doing. Next thing that got people really upset. We have clients, let's see, we have 435 clients across 42 states, and we're about 500 employees strong. All our armed and unarmed agents, primarily unarmed agents, are on our payroll, W2 employees. The security industry is broken. 100%-400% employee turnover rates.
They can't keep the people on the job. A security guard that's standing outside this hotel who joined here maybe in January, by April, he's not our post anymore. Right? They can't keep the people on the job. Why are we at 6% instead of 400%? We recruit very carefully. We train properly. We pay all their health benefits. We give them stock options in Knightscope. Yes, that security guard actually has a financial interest in the success of the company. Then we're going to outfit them with this crazy technology that takes some of the capabilities of our robots and puts it on their person. Right? Then there are going to be our mission intelligence people, our remote monitoring folks are in their ear. There's no one going to jump one of our guys because we have eyes everywhere.
What this has is a forward-facing camera, a rear-facing camera, a thermal scanner, so we can do persistent fire watch. It can do facial recognition, license plate recognition, and also detect a mobile device within 500 ft of the machine. There's no way you're going to sneak up on one of us. A lot of people got really cranky about this because it's proprietary to Knightscope. We are not selling this to anyone else. That goes with that vehicle that you just saw. Then people get really frustrated because what if you're able to take video gaming technology, digital twin capabilities, and provide a mission intelligence analyst a view that no one in the world has, with no blind spots, to actually remotely monitor a location with new capabilities? We said we're going to go build this crazy piece of software. We've built it.
We've now deployed it this week at our own facility. Who wants to see a demo, real Oh, yeah? Okay, good. I'm going to walk you through this demo of the capabilities of Signals. Remember all the pieces of hardware of our arsenal, all the cameras, the LiDARs, the vehicles, the humans, the stationary devices, all that is to feed data into this twin so that we can keep it alive. You have a digital twin in your pocket. It's just very old. Your Google Maps, your Apple Maps. Until that Google Map dude walks around your facility or outside your facility, you have stale data. What we have, because we have the entire stack of hardware, software, and humans, we can keep this alive. No one can do this. That gets this to be the glue that ties the entire thing together.
What you're looking at is a 3D model of our facility in Silicon Valley, in Sunnyvale. You're going to be able to literally see all the assets that feed data into this machine. The arsenal on the left is our portfolio of products. I'm not going to go through the entire list of products there, but we have stationary devices. We have mobile robots at 3 mp h, at 10 mp h. We have new emergency blue light capabilities. We're integrating camera technology that exists already that a client might already invested in that infrastructure. LiDAR technology, not for autonomy, LiDAR for detecting capability. The whole business here is to deter the negative activity, to detect the activity, and then to respond. Right? 92% of the alerts in the security industry are false. 92% of alerts are false.
We have 100 million cameras in the U.S., literally 2% of them are remotely monitored, and the rest of them are just spilling out junk. What you're able to see here is all the assets that are there for you to be able to have a real time of the facility that we're securing. Think of critical infrastructure, manufacturing facilities, data centers, water treatment plants, and the list goes on and on. Solar farms, wind farms, actual farms, ranches, et cetera. Now this crazy view is you're able to control, either autonomously move that virtual camera, or you can actually do an eye level patrol as if you were physically there. Again, the only way you're going to keep that digital model up to speed is you have armed and unarmed agents that's spitting out a massive amount of data.
You have robots spitting out a massive amount of data, and then you've got all that integrated in one crazy sensor fusion to give you an ability to have a live digital twin of your facility that no one else can replicate. That data gets really interesting in order to remotely monitor. Imagine if you got a Ring doorbell camera, right? You're looking at the app, and you're like, "I can't see what's to the left of that, to the right of that. I don't know what's behind her. I don't know what's in front of her, actually. But sure, I'm going to remotely monitor a location." Good luck. That's not going to work. That's like an idiotic assignment. This gives you an incredible field of view. I'm going to give two incidents that we're going to walk through. Up on the right there's something approaching the perimeter.
We can detect a mobile device within 500 ft of the machine. The LiDAR didn't see it, the camera didn't see it, the human didn't see it, the robot didn't see it, but we detected that there's something incoming. I don't know what it is. No, that's you. The camera and the light on the building turned on like, "Hey, I don't know if that's a guy. I think it's a person." The camera is now going to call the robot over and say, "Robot, come to this longitude and latitude, run a face detection, a people detection, thermal scan, and mobile detection, and tell me your probability of success that that is a human." Robot comes back to the camera and says, "Yep, 99%." Now I'm going to do a talk-down. What is a talk-down?
Sometimes when you tell a person, "Hey, it's 2:22 in the morning. You're trespassing. Please vacate the premises. We're calling the authorities." Sometimes the people leave, let's say half the time. That's what just happened. What this does is combine AI agents, that digital twin, the robot, and a bunch of orchestration software so you're able to autonomously auto-resolve ongoing issues that should never have been escalated to a human in the first place. Now the guy vacated the premise and issue resolved. You got an all clear, everything's green. You're good to go. Looks like a crazy video game, because it is. Next incident. Let's say someone pulls into the parking lot driving too fast. We've got a solar-powered tower that says, "Hey, there's a vehicle that came in way above the speed limit. I've got the license plate. I got the facial recognition.
That person's been on the lookout." I've got a live stream at the bottom right for the analyst to review. There's no way we're auto resolving this. I need boots on the ground. Our mission intelligence command folks, they've got a view from the robot at the lower right. They're able then to escalate this to our staff. Our augmented security agents to be boots on the ground. That little ASA that's coming up over here, that is a person that's going to physically go confront the intruder or speeder and try to de-escalate the situation. That's an interesting way if you had the entire stack of not having an escalation model that's binary. I had a city manager tell me this. "Our escalation model in the city is binary." What the hell do you mean by that?
Zero or one, you either don't call the cops or call the cops. That's not the way to do this. You've got all kinds of privacy issues going on with license plate cameras and that sort of thing. What you need is an escalation model. First layer, can you do everything autonomously? If the AI agents and the robots get confused or don't know what to do, escalate it to the remote command human, and then the remote command human, then and only then after it's been reviewed, do you dispatch boots on the ground. The last call is to law enforcement. That's level seven.
The chief of police turns to me and goes, "Ah, so when you guys call, I should actually respond." Because the 19,000 law enforcement agencies are 30%-40% understaffed and may or may not show up and cannot take on an unwanted call. That is how we clear things out. One of the craziest things in this $230 billion TAM is proof of work. Did you show up to your shift? Were you texting? Were you sleeping? Were you eating? Did you do your rounds? Can you prove that you were there? Pretty much no. This is the only industry that I know where an officer of the company is literally signing invoices, $1 million, $2 million, $50 million, $100 million, can't prove that the work was done. It is insanity. What did I tell my team?
Hey, guys, this is really cool orchestration software that's proprietary to us and us only. Can you pull the log files for us?" "Bill, what do you need the log files for?" "Just print the log file." There are two ways to print the log file. One, the reason that turned black and white is you're going back in time. I am not an attorney, but I can imagine sometime in the future, "Yes, Your Honor. We have all the evidence to prosecute to the fullest extent of the law. We can recreate the scene in three dimensions. Which angle would please the court?" Complete game over. You can do this in three dimensions, and up on the left-hand side is what's getting populated, and you can kind of replay the entire scene. For sports fans, it's literally a replay of the play of what happened on the field.
But you can do that in three dimensions, and then you can move the camera around any which way. So there are no ifs, ands, or buts of what happened. The other thing that you can do is actually just print out the timeline of everything that happened down to the millisecond. Give me the longitude, latitude, timestamp, which asset was involved, the AI decision log, why did you make certain decisions, and everything is there, fully exportable, automatically created and issued. Not the guy minimum wage out in the parking lot with a crayon or a number two pencil trying to write a report of what happened. This gives you an unbelievable level of capability that no one else can deliver for this industry. This is the reason why we're really excited, and why a lot of people are really cranky with us.
The next number I want you to think about is triple. This week we announced triple-digit revenue growth three quarters in a row. Triple-digit revenue growth three quarters in a row. Interesting times in the marketplace. I guess we get punished for continuing to deliver results. But what we've committed, the CFO and I, every 10-Q and every 10-K going forward is going to get better and better and better. The way we're going to teach the market that this brand-new category that's never existed is by delivering results. Again, we're a managed service provider. The category is all new. This is not a product. This is a service you commission. The TAM is about $230 billion, focused on the U.S. and the U.S. only. This is a recurring revenue business model for a recurring societal problem. We do not have a market risk.
The market for crime is not going to collapse tomorrow. This is not station wagons or phones or whatever else. This is a recurring societal problem. The moat gets really interesting because the more data we get, the better we can make the service. The better we can make the service, likely the more clients we get. Then we just keep building this crazy amount of capability, for us to deliver incredible service at lower and lower costs. Why are we doing this? Besides wanting to make the U.S. the safest country in the world, the chief security officers of the United States of America have been dealing with an absolute cluster, which is on the screen right now. Eight to 12 different vendors. Investigation team doesn't talk to the remote monitoring team.
The data coming from the remote monitoring team doesn't go to the agent out in the field. The agent out in the field doesn't talk to the robot. It's just a big mess. All of these vendors are really good at shipping what? Invoices on time with no accountability. In our case, we ship one invoice. We're one entity that's accountable, and we have the audit trail to prove it. There is not one chief security officer in this nation that I know that's going to say, "No, I don't want a better team at lower cost for better service." That's not a thing.
This goes back to the seven layers of escalating things properly, so that you don't have any privacy issues where you've got multiple agents, AI agents, multiple humans involved, so you're not a law enforcement officer with access to license plates that can stalk his ex-girlfriend of what she'd been doing the last 300 days. You're not able to do that. Second, everything in process to be federal grade for the entire country, state, local, federal, plus our private sector clients. So, DoD Impact Level five cybersecurity and FedRAMP High. So now you have high confidence in what's being done with the data, and it's federal-grade cybersecurity for the entire nation. Why do we think we're going to win this? I often say, and you probably read it, every battle or war is won or lost before you even fight it. We already won this.
We just have to execute. I don't have a market risk. I'm a former engineer. Everything can always be made better. We don't have a technical risk. We only have to focus on execution, and the best way to do that is to hire brilliant people. We're 500 strong now, and we're really excited. Basically, what this says is the three largest guarding companies in the United States, all three about to be all foreign owned, which is going to be another discussion that I'm going to have with the federal government, is how can you have the three largest guarding companies in the United States of America be owned by foreigners, who now have access to all our blueprints, all our critical infrastructure. They know every weakness and can expose every issue. That is not in the national security interest of the United States of America.
Us being full-blooded American, NASDAQ listed, with proper governance and audit will certainly be to our benefit. Basically, what this says is we have the entire stack accountable under one company. This, with six minutes left, is where I ask for your vote. I am the founder of the company. I actually believe we can make the U.S. the safest country in the world. Just pause for a moment for reality, and let us say we pull it off. Talk to me about a municipality's budget if we did that. Talk to me about property values. Talk to me about insurance rates. What about the viability of someone's local business, the safety of your family, frankly, all of your portfolios? What dramatically changes everything for everyone. We just need to execute.
I am asking for your support, not just you, not just you, not just you. I am asking for everybody's support in this entire room because this is what it is going to take to build this company. With that, happy to take your questions. Yes, sir. We have been operating in the field fully autonomously for four million hours since May, if you are a Star Wars fan, May 4th, 2015. We have years and years of data. We have clients that have renewed with us for five, six, seven, eight, nine years in a row. You are not going to get a client to pay you full price for a decade if you are not delivering results. Not every site is the same. Some sites, I will be super frank with you, they are negligent, and you need to increase the budget.
Some sites, dude, what are you doing? You have 16 officers running 24/7 trying to secure four vacant buildings, and you are paying one of these three guys $2 million a year? I think we can fix that for you, right? Some of it is going to be dramatically 50% improvement, and then somewhere in between. What I want to do is for us to be a trusted advisor, we have 435 clients across the U.S., is for us to go line by line. Look, not by client, not by vertical, by address. What is the problem that you have at that location? How can I reduce your cost? How can I improve the service? It is the painful way, but it is the right way to do this. Yes, we have data.
One of the annoyance for the media and for investors is please tell us all your clients and all that. We are not doing that. For what we do, we are just not sharing that. Yes, ma'am. We do anything from a retail, small, tiny location, small hotel, to a 500,000 sq ft critical infrastructure. It highly depends. I am not trying to be cute here. It highly depends on the location. Maybe a few hundred thousand square feet per robot. But if you tell me, "I have a lot of incidents and I have a lot of traffic," that answer might change. Right? Yes, sir. Oh, yeah. I have only got two minutes here, but I will try to do the short version. Everyone keeps saying, "You guys are burning too much cash." I am wildly offended. We are not burning cash.
We are investing in building technologies and capabilities that no one in the industry has, and it will be the best dollars we ever invest. We typically don't like to do deals, whatever crazy suicidal warrants and stupidity that's out there. We primarily rely on our ATM. Most people don't like that, but it's the lowest cost of capital for us that's non-toxic and clean. And we pull down capital when we need it, not in massive tranches. You can look at our financials. Yes, are we going to need more capital to get to breakeven? Yes, absolutely. But I'm not playing games here. We've got line of sight to get to $1 billion of annual recurring revenue, and we can pass the red-fac e test. I'm going to do whatever it takes to get the company funded fully so we can go execute. Yes, sir.
Contract length. Typically, the security industry runs on year-long contracts, more or less. Gross margins, we're landing with humans and then expanding with technology. If you're operating okay, you're probably 10% or 20% just humans. You do it really well on the technology side, you should be in the 50%, 60% gross margins. And so, what we need to do is get our foot in the door and then re-architect a solution so we can improve the margins and reduce the cost for clients. Yes, sir. We do have thermal cameras. LiDAR, we use on the buildings for detecting of vehicles and humans. But we also use LiDAR for navigation, and all our devices have thermal. That's not how we do a technical conversation outside. That's not how that works. Yes, sir.
Is there a possible [inaudible].
What we try to do is a subscription instead of a Chinese menu like, "Hey, do you want this LiDAR? Do you want this?" Here's all your problems, here's one price. And that could be on a monthly subscription, or it can be a one-time deployment fee and then a monitoring and support fee. But the end is, the answer is no. It isn't. In some form or fashion, but yes. Oh, she told me to stop. I can hang out for a little bit. Thank you everyone for attending, especially on late notice.