What we talked about
Michael Green explains why he co-founded Athena Security and how a real, heartbreaking incident in 2023 reinforced the urgency of their mission: saving lives by detecting and denying weapons:and sometimes dangerous people:before they enter facilities. He shares how targeted violence can slip through gaps in coverage and how that reality pushed his team to deploy faster and more widely.
Show notes
When a nurse was killed by an ex-boyfriend with a restraining order at a facility in 2023, Michael Green’s team at Athena Security realized their system could have caught that person twice over, once through ID inspection and once through facial recognition, if only it had been deployed there. That incident transformed a product roadmap into a personal mission.
What we covered
- Athena’s visitor management system performs multi-point ID matching using name, date of birth, and optional facial recognition to detect people on trespass lists or protective order lists before they ever enter a building, two independent chances to stop a known threat.
- The technology uses multi-frequency electromagnetic detection to distinguish between different metals, alloys, shapes, and sizes, allowing facilities to define their own threat targets, from handguns and large knives down to small blades and OC spray, and tune the system accordingly.
- False positives and alarm fatigue are the primary ways security technology fails in practice, not technical limitations. The more contraband categories a facility tries to screen for, the more secondary searches officers must perform, and human consistency degrades with repetition.
- Athena uses AI to enforce officer accountability, tracking whether security staff are physically at the machine, logging why each alarm was triggered, and building an audit trail that shows every person was cleared through a functioning, calibrated detector.
- Social engineering through familiarity is a real vulnerability: a repeat visitor who becomes friendly with guards can slip through without inspection, as illustrated by the story of a smuggler who walked through a military checkpoint daily until the personnel rotated. Technology enforces consistent checks regardless of familiarity.
- Green described the future as “tunnels of truth”, integrated lanes where weapon screening, X-ray, visitor management, access control, and video systems converge into a single operational view, with digital identity verified on arrival so the physical screening becomes seamless and nearly invisible to the person walking through.
About Michael Green
Michael Green is the CEO and co-founder of Athena Security, a company focused on keeping weapons out of buildings through AI-assisted object and behavior recognition. Athena provides a full security operations model that combines hardware, software, training, and ongoing service support for facilities ranging from schools and enterprises to events and venues.
- LinkedIn: https://www.linkedin.com/in/michaelleonardgreen
- Website: https://Athena-Security.com
Episode 97 of the PreVetted Podcast.
Full transcript
Federico Ramallo (00:01) Welcome back to the pre-Vered podcast where we spotlight extraordinary people and remarkable talent reshaping our world. Today we’re joined by Michael Green. He’s a CEO and co-founder of Athena Security. Michael is on a mission to keep weapons out of buildings and make public spaces safer through technology. At Athena, his team provides threat alarms using object and behavior recognition for threats like guns,
knives and fights and supports organizations with a full practical security operation model hardware, software, training and service designed for long-term sustainment. Michael, welcome to the show.
Michael Green (00:44) Rico thanks for having me.
Federico Ramallo (00:46) Awesome, awesome. So we’re going to talk a lot about what you’re doing at Athena. Tell us first about your mission, keeping weapons out of building. How did you come up with that mission?
Michael Green (00:58) Sure, we’re on a mission to save lives and it’s a huge mission that helped protect by detecting and denying both people as well as weapons in the facilities. And this really hit home when we were doing our first installs in major systems in 2023. And I was able to go on site and major customer and it was really high profile and aim everyone would know.
in the industry and after we set up the customer we did Google news alert and we actually made it company wide and soon after we had an alert come through the say a nurse was killed shot and killed by an ex-boyfriend and we read in the article that it was someone who had a restraining order in addition and so it’s quite a heart-sickening story of a known person a known assailant that targeted a specific person right
So it’s targeted violence. Anyway, my heart sank and in fact, this is why we developed this kind of system. And so this is not a building we’re in yet as we investigated and the gun did not go through our lane of detection, but this highlighted why our visitor management systems there that detect the deny this kind of Bolo and our Apollo 500 would have caught this kind of contraband it’s designed to and.
I feel heart, heart, you know, very sad, very heart struck that we weren’t able to help the specific person, but it actually lit a fire under our whole company for more widely deploying this technology and more of our systems, more places.
Federico Ramallo (02:23) Right is that feeling of, you you could have prevented it.
With the technology in place, you could have prevented that situation.
Michael Green (02:30) That’s right, yeah.
Two different chances to catch that person. One through just, if they put their ID into our cradle, it would have detected and denied based on name, date of birth. So we do multi-point matching and inspection of IDs. We also can do facial recognition if we’re allowed by that facility. So they can turn off or on that kind of feature.
And so that gives us two different ways by either the information on a license or that facial rec match to find that person on a trespass list or on a protective order list. Then in addition, we could have kept out that contraband by running everyone through screening. So a sound screening program would set a threat target in order to find that kind of knife or blade.
from pocket knife and bigger guns, handgun, shoulder fire gun, so lots of different kinds of guns. It’s meant to cap alert on.
to help keep out and so we’d be able to define all those different thread items, set the detector accordingly, and our systems, it’s made to keep that out. Of course it needs humans behind it driving secondary searches, making sure that they’re doing great secondary searches consistently to find the contraband, locating it, removing it from the environment. But that’s why we’re here partnering with those organizations so that we can build those kind of systems.
Federico Ramallo (03:43) Right, right. The way that I usually see those type of security system work and fail is when they fail when they have a of false positives, right? And when it requires the security organization to do something extra that what they’re doing, because they already have their hands full, right? So if you ask a security guard to do this additional check and do this additional, you know,
Eventually, unless it’s strictly necessary, they kind of stop doing it. So I like what I was reading about Athena because it doesn’t require user’s input. It can work automatically and autonomously, right?
Michael Green (04:21) Yeah, so the smaller threat target people want to find, the more alarms there will be, just necessarily. So if you’re only screening for handguns and large knives, then we can have minimal alerts, but if it’s going to actually screen down towards small blades, then it will necessarily have lots of secondary searches. But you’re right, the more secondary searches there are, the greater chance there are for a human error in secondary screening procedure.
Federico Ramallo (04:26) Right.
Right,
yeah. And the lower the quality, the more secondary searches you have to do because it’s a repeatable job and it’s a thankless job at the end of the day, right, for the security guards, yeah.
Michael Green (04:59) That’s right, so more professional screening
staff can do a more consistent job because they’re just more focused and can maintain a high level of performance over time. that’s what we’re here to support.
Federico Ramallo (05:10) Amazing, So, when we’re talking about Athena security, you mentioned that there are different objects they can detect and different potential threats that it could identify. Can you tell us a bit more about that?
Michael Green (05:25) Sure, so we’re using multi-frequency active wave.
electronic magnetic, electromagnetic detection. I know that’s a mouthful. I even get a little tongue tied saying it, but it is metallic based and the whole not the whole invention is around doing better filtering on different kinds of metals, different kinds of alloys based on shave size, size, mass in order to filter out what kind of things are coming through. And so it’s really up to customers to find what they’re trying to keep out. And as I mentioned, events, venues are all trying
keep out weapons and very large knives. So that’s what we call high throughput screening because you can still keep a cell phone in your pocket and go through and the system can do a pretty good job at telling the difference. But it’s really in the smaller things that then we can start separating out from the competition how we can do a better job finding edged weapons and blades and still find more contraband with fewer alarms than our leading competitors.
Federico Ramallo (06:21) Right. So you can detect the shape of the objects and based on that, you can define what type of object it is.
Michael Green (06:28) Yeah, these are metallic items.
Federico Ramallo (06:31) Right, right, interesting.
Michael Green (06:32) So
iron based as well as non-iron, so aluminum, sleds, all the different kinds of metals.
Federico Ramallo (06:38) Right, because it doesn’t depend on magnet, it’s not magnetic detection, it’s metal detection.
Michael Green (06:44) Correct. We’re using a very light electronic wave, the excite metal coming through.
Federico Ramallo (06:51) Right. And what happens when a potential threat is identified? I guess it depends on the policy of each building, right?
Michael Green (06:58) That’s right. Most of our customers proceed the secondary screening with either a hand wand. That would be most common in order to start with an isolated screening in a specific area or region of interest. So that’s how most of our customers handle it.
Federico Ramallo (07:11) Amazing. So Athena works across public spaces like schools and malls and also enterprises, right? What changes the most across these different environments?
Michael Green (07:23) I think the biggest difference is between like an event where it’s a lot of focused attention. So if you’re, if you think of going to the last NFL game or football or, you know, we call it soccer, any of those events, you’re going to overstaff your security team and it’s going to rely more on those officers. Then I think,
Let me rephrase that. When you go to events, they can afford to overstaff it because it’s a tight time window. When you’re running a system that’s 24-7, so any facility, the longer hours you have to run secondary screening, that officer or set of officers is going to get just more fatigued because they do it more routinely all day long, lots of days in a row. And so there’s just more fatigue that comes with running 24-7 longer.
Federico Ramallo (07:48) Right.
Michael Green (08:09) on duty times. So many of our customers do things like rotate posts in order to keep their officers fresh. But that’s where technology really needs the help.
officers stay diligent, build accountability into the program. So we help officers resolve alerts so each alarm can actually be resolved. So officers should understand why the machine went off. They can mark wheelchair walker, know, things like that that set the machine off versus if it is in fact a knife.
Again, most of our customers are finding edged blades as most of the contraband coming through. Other things that we find a lot are things like OC spray, like bear spray, pepper spray, and different concentrations. So things that are legal and people are not allowed or prohibited that bring into a facility are what we most commonly find. Work blades, probably number one. So that’s what we’re trying to…
help our customers prevent from coming in because that’s their, again, their threat target that, their threat targeting that decides if they want to try and keep that out or not. But really the biggest difference is if you have a concentrated effort of human staffing at a event or venue versus 24-7, you’re going to have to more thinly staff that necessarily.
So those are the biggest differences.
Federico Ramallo (09:17) Right, right.
The other thing I’ve is this concept of repeated, you mentioned repeated task, but also I’ve seen repeated people, right? You see the same people over and over and you kind of make, I don’t know if friend is a word, but you know, a friend relationship. So I had this story where there was a military checkpoint in Mexico and there was this little old man just.
walking through, you know, and they checked them for the first time, but every day was passing through and nobody, know, they wouldn’t check them, right? And then something happened that they changed the people on the checkpoint and then they check again the little old man and turns out that he was smuggling drugs through the checkpoint because he was a repeat, you know, he was walking by so frequently that
he became friendly or well-known for everybody, right? So they wouldn’t check them again, right? So he kind of was able to penetrate the checkpoint using that social engineering.
Michael Green (10:18) That’s right. Habituate, become friends with, say hi. Pretend like you’re not an alarming person, become friends with them, be on a first day basis.
Federico Ramallo (10:21) Right. Right.
Which doesn’t happen on a big sport event because it’s a lot of people and nobody knows anybody, Their alert is higher,
Michael Green (10:38) Yeah, I would say social engineering is one of the, for business process outsourcing for a whole set of customers that are new to using any kind of metallic threat screening coming on board because they’re actually afraid of electronics coming in.
So we’ve been able to also build a better mobile phone detector with the same technology in order to keep out smart devices, smart glasses. More and more people have, you know, it’s millions of smart glasses that are being sold every year now. And so all of sudden, companies don’t want those allowed in because that’s how they can steal data as a new form of contraband, a new way of socially hacking a company.
Federico Ramallo (11:06) Right.
Right, right. And sometimes you want an event with no cameras to have some privacy or in concerts, I remember about artists that they don’t want cameras there so people can enjoy, you know, like enjoying life like in the 90s, where you would live the moments. That was kind of the idea of this artist. I think it was a rock star. I was complaining because everybody would pull their phone and they basically miss the experience, right?
Michael Green (11:40) Exactly.
Federico Ramallo (11:41) And what particular challenges have you found with schools?
Michael Green (11:45) gosh, humans are ingenious and kids are very creative. So schools need to do a lot before they’re even ready for a technology like ours because they need to secure the perimeter first, which is incredibly hard because they’re large campuses. So schools have a ton of challenges to do great weapons screening programs and keep weapons out because kids are so ingenious and the facilities are so large.
Federico Ramallo (12:07) Right, right. Unlike an enterprise building where they have one entrance, one exit, or very minimal attack surface, right?
So you emphasize accuracy, speed and convenience. What trade-offs do you have to make to get all three working together?
Michael Green (12:24) man, that’s like going to a developer saying like, I want this app, I want it to develop now, I want it to work great, and I don’t want to pay much. So you can’t have everything. I’m sorry.
Federico Ramallo (12:33) Right.
So how do you balance these three attributes?
Michael Green (12:39) Yeah, so it’s a necessary trade-off. So we
talk about that all the time, but it is necessary friction.
So there’s some necessary friction when you enter a passcode to get into your computer that use a key to get into your front door. So there’s some necessary friction for doing weapon screening. And we’re just here to offer the best that our engineering team can deliver right now in order to help make it the most speedy expedited process possible to help reduce alarm so that there’s less alarm fatigue from officers and then wrap all this up with a great deep analytics package and lots of other adjacent
products in order to be able to like for instance use x-ray right next to it so that you can do a better screen on property searches. We use computer vision so when you put things like a backpack through an x-ray we can find contraband faster with AI in the x-ray so it’s a great opportunity to be a workforce multiplier. So that’s a great example of deeper screening faster.
but it does cost a little more money then in order to add an x-ray as well as then you have the space for it.
Federico Ramallo (13:32) Right.
because the machines are bigger, right?
Michael Green (13:40) Yep, another three feet wide by nine feet is probably what you need to add an X-ray. Minimal. Yeah, minimal. And that’s in the entryway too. Where space is a premium, yeah. ⁓
Federico Ramallo (13:44) Wow, that’s, yeah.
Right, is
a critical real estate there, yeah.
Michael Green (13:52) Yeah, with safe egress, so you have lots of other constraints around safe egress, life safety issues.
Federico Ramallo (13:58) Right, right. So sometimes the trade off is based on the location and the building and the requirements, right?
So what are the most common reasons security technology fails in the real world, even when the tech is good?
Michael Green (14:12) Well, I’d have to defer the other experts on what exactly the most common failure points are.
What do you think they are?
Federico Ramallo (14:20) I mean, for me, think it’s we were kind of talking about this, right? People, human error, fatigue, right? Having to repeat things, social engineering. Because I think most of the technology is good enough, right? And then from there you have great, right? The technology made a lot of improvements on false positives. But then…
if you have a friendly guard that makes, you know, that was able to skip the checks, right?
Michael Green (14:44) Yeah.
Yeah. My answer is pretty boring. Consistent discipline process.
Federico Ramallo (14:51) Yeah, and it’s this idea of the boredom that makes people to lower their guard, right? Because they repeat.
Michael Green (15:02) That’s right. So we use AI to
actually keep officers accountable. So we can do things like we call it officer check-in so we can understand if the officer is at the machine or not. So it’s a feature we build for a customer to make sure the officer is actually there. In addition, building in accountability with alarm resolution. That’s why I mentioned it earlier, to make sure that officers are actually doing a secondary screen and then they’re responsible for inputting why the machine actually generated the alarm.
Federico Ramallo (15:20) Right.
I see. I see.
Michael Green (15:28) Yeah, we use
AI for a couple other things that can be employed to make sure everyone went through the machine. So we have reporting with a audit trail on everyone that was cleared through the device, as well as then we can make sure the technology was actually calibrated, powered, connected. So we make sure the audit the technology, the make sure that’s good to go. And then also,
roll all this up with deep analytics. So everyone went through the detector. The detector was on work and calibrated and officers were physically at the machine and doing their job for secondary screening. So we helped build a more accountable process.
Federico Ramallo (16:02) Right, Yeah, tracking the behavior of the security guard, I think that’s game changer.
Michael Green (16:09) Yeah, AI for workforce.
Federico Ramallo (16:11) Right, right, and supervision of behavior, right? Because before it becomes an issue, then you can detect risky situations, right? Or when they’re lowering the guards, right?
Amazing.
So you mentioned choosing the best plane of detection and sometimes adding X-ray. How do you decide what is the right setup for a specific building? What is your process?
Michael Green (16:33) you
Sure, we lean on set of architects, consultants, engineers, and specialists to help provide that kind of advice. And we’re here to provide the best technology. So we’re more experts in the AI, in the software and hardware layers, that understand capabilities. And so we help that community understand capabilities so that they can use it most properly.
Federico Ramallo (16:58) Amazing.
So you’ve said that this approach can reduce staffing and labor in some cases. Where do you most often see those savings actually happen?
Michael Green (17:09) Sure, so long term, most of our customers are able to save some labor by reducing some of the staffing need over time for when there’s very low throughput.
Federico Ramallo (17:20) Okay, okay, so when when you have a building with high throughput and low throughput then on those times is when they can
Michael Green (17:25) Yeah, so if you
have three people staffed at a machine at night and it’s only two people coming per hour, that’s an opportunity, probably the same staff.
Federico Ramallo (17:34) Right. Right. Interesting. Yeah. And also you mentioned that the platform allows you to supervise behavior. you can reduce bandwidth and supervision as well. Interesting. So how do you think about privacy, transparency, and trust when deploying detection systems in everyday public spaces?
Michael Green (17:44) That’s right.
So we’re not using any of our technologies in rights of way in what you would consider like public space. We haven’t yet anyway. usually it’s a private property owner. And in the United States, that’s who can control. And they put up the signs that say, hey, you’re under surveillance when you come on our property.
Federico Ramallo (18:12) Right. So they can decide who let them in or who don’t. It’s a matter of a policy of the owner of building.
Michael Green (18:20) That’s right, yeah.
Federico Ramallo (18:21) What metrics do you use to prove safety outcomes to customers?
Michael Green (18:26) metrics to prove safety.
Federico Ramallo (18:28) Bright safety outcomes.
Michael Green (18:30) So, something didn’t happen, you’re right, is very difficult. And I think that’s what you’re getting at, So when you study large numbers, you can start looking at incidences, incidences in specific areas of buildings. And so, if you have to look at, you can look at high level numbers to look at how you can change outcomes. So it’s only in large numbers that you can start looking at case studies, where it’s a business case around reducing
the occurrence of violence and mitigating that catastrophic risk. So we have a white paper on it. It’s like 22 pages. Happy to send it to you.
Federico Ramallo (19:00) Right.
interesting. Yes, yes, I would love to read it. Yeah, the it’s it’s a story of who’s a better captain, the one that went through the storm and survived or the one that avoided entirely. Right. But everybody.
Michael Green (19:05) Yeah.
You’d rather, yeah,
mean, you’d rather not need it, but if you need it, you’d have it than need it and
Federico Ramallo (19:25) But then everybody will complain that they arrived late, right? But they didn’t see the storm that they avoided, right? Which is kind of what you were talking about. How can you prove what didn’t happen, right?
So what do you think the next three or five years would look like for threat detection in public spaces? And what do you hope changes most?
Michael Green (19:47) we see a huge convergence in technologies. So right now, more of our customers are asking for more. So like our walkthrough weapon screening lane integrated with X-Ray, integrated with visitor management, integrated with access control, integrated with video management systems. So there’s this convergence now of all systems being connected so that directors of security can see all of it through one operational, you know, a G-sock or a sock. So they want to
Have all alerts and events roll up to one place so that they can understand and manage all systems and all people all processes in one spot for faster investigation faster research on any problems as well as then Being able to detect and deny people from coming in that shouldn’t be there making sure contraband doesn’t come through through the planet detection or x-ray and the physics of detection are going to continue to
be the challenge for engineers in order to make better detection units. So all that’s happening, think additional wavelengths are going to start augmenting that metal.
That’s been the traditional way of finding contraband that gives us the opportunity that you use object detection to find things So I think in the next five years that’s gonna start being commercially available as well as that Convergence that I mentioned so you’re gonna go through like a tunnel of truth for the airport You’re gonna have some kind of digital ticket that’s on your phone. You’re gonna be able to walk up They’ll know who you are while you’re there what flight you’re about to catch everything’s gonna be cleared probably through clear and Connected you go through your physical screening and then you board your plane
So more of that is going to be digitized through these kinds of tunnels of truth concepts where you’re not gonna It’s not going to be so in your face and so manual. So that’s what I think the future is
Federico Ramallo (21:25) Amazing, Michael, thank you very much for joining us today. Any final remarks before we wrap it up?
Michael Green (21:32) Hey, it’s my pleasure and we just want to be a great provider. And if you have anyone out there in your user base that’s listening to this, we’re always looking for great engineers, great advocates for what we do for the technology, as well as just great security practitioners that want to join our family of.
installers, resellers, so that you can help keep weapons out of facilities and people that shouldn’t be there.
Federico Ramallo (21:58) Amazing, amazing. And we’ll put the links in the description below so people can reach out. Thank you, Michael.
Michael Green (22:01) please do.
Thank you so much. Have a great day. Thanks.
Federico Ramallo (22:07) You too.