Innovating Leadership:
Co-Creating Our Future

Resources:

Learn more about the work of Brad’s company on their website.

Our host Maureen Metcalf posts a newsletter every week on LinkedIn. You can subscribe here.

Maureen’s latest book is Innovative Leadership & Followership in the Age of AI. You’ll find details about it at https://bit.ly/LeaderInAI, or check out the Kindle version at https://amzn.to/44buVz8. The audiobook version is now available at  https://amzn.to/4dTCleZ.

Her other 10 books are available on Amazon here.

Other episodes you’ll enjoy:

– Fear Less, Shine More: Build Confidence in Your Leadership (and Life) with Tonjia Coverdale

– To Stop a Tyrant: The Power of Followers with Ira Chaleff

– Forget Power: How to Lead in a World of Chaos & Complexity with Michelle Harrison

Guest(s):

Guest(s) Bio:

Brad Diskin was CEO of UROK Learning Institute for 20 years. The firm specialized in teaching students with learning deficits to read, write and do math. The firm taught students to read nationwide utilizing the company’s proprietary reading program Literacy Links. One of the tools became the genesis of Sound Genetics Inc. The company received a United States Patent on August 22, 2023 SYSTEMS AND METHODS FOR PRE – FILTERING AUDIO CONTENT BASED ON PROMINENCE OF FREQUENCY CONTENT.

Transcript
(auto-generated)

Maureen: Welcome to Innovating Leadership Co-Creating Our Future. I’m your host, Maureen Metcalf, founder and CEO of the Innovative Leadership Institute, where we help leaders be future ready. Today we’ll explore the intersection of technology, leadership, and transformation by diving into the fascinating frontier.

Acoustic intelligence with Brad Diskin, CEO of SGIA company pioneering the use of sound as a new signal layer in AI systems. SGIs work is redefining how machines perceive the world and it has powerful in, in. Beep SGIs work is redefining how machines perceive the world, and it has powerful implications for security, fraud prevention, and operational [00:01:00] resilience.

So that’s quite a robust technology. Brad, let’s start with the big idea. SGI isn’t just another sensor company, it’s creating a new sense. For machines. So let’s delve into what behavioral audio intelligence means and why sound is such a critical missing piece in ai, and I wanna say and why leaders care or should care once

Brad: Okay, well, well, let me start by saying that it, the, the, the idea was, uh, not actually a big idea. We came across this technology through the intersection of EdTech and, uh, and tech. And so what you’re looking at behind me is a, uh, a demonstration of our technology. This is the visualization of the sound spectrometer.

So you’re looking at a tool. That the world has never seen before. It is actually going so far deeply down into the sound pro profile [00:02:00] of my voice. We’re seeing 34 different dimensions of my voice on the lower screen, 10 different dimensions of my voice on the upper screen. So this is a way to see inside of sound that is never been done before.

So we started this, uh, this nine years ago. And we have, uh, tried to do many different types of things. We did delve into Phony id, which was the first one that we did. We did song id and then we were discussing with Silicon Valley Bank in California, where would the world really benefit from? Something from SGI.

They point us toward voice authentication, deepfake detection. So what we’ve now developed is a deepfake detection tool that is now capable of understanding the difference between a human and a speaker or a generative voice. That is the first, uh, tool that we put in front of any system we can integrate with anything in the world.

Uh, it is, uh, machine agnostic. Uh, so the [00:03:00] deepfake detector would be the guard at the gate that would take a look and see if the signal coming into the system is a true human being. And the reason we can do that is that the machine behind me can define what a vocal cord sounds like. We know the difference between a vocal cord and a generative signal Coming from a speaker, we have most all of the speaker parameters into the system.

So as the audio signal comes in, we know human. Or is this a speaker? If this is a speaker, it stops that particular audio signal from coming in and no longer is the deepfake coming in. And so we go from there to authentication. Now we can go to authentication. We’ve created an authentication tool utilizing the same platform.

Where we can now, uh, be able to do a voice authentication of a human being. Every human has a unique voice. Every human has. This is their fingerprint of their voice. This we can create an MRI, the DNA of their voice and create the [00:04:00] voice signature quickly, efficiently, cost-effectively, and uh, uh, and useful.

And so we see a wide range of applications here for both deepfake detection and voice authentication in the same package.

Maureen: So I’m thinking of. A wide range of ways this would be helpful for me as a business leader and also some ways that I don’t like it. Like I, I have my own not deep fake. We have, I have a digital twin and I love not having to record videos. I love being able, and we label them Maureen’s digital twin. I’m not trying to inauthentic, but I, I do appreciate having the opportunity to. Not have to record videos.

Brad: Yeah, and I, and I see that because, uh, we’re not trying to stop people from using their digital twin to be able to get their work done. What we’ll be able to do is to know if you’re using your [00:05:00] digital twin or if you’re using yourself. So your digital twin will now have a guard. The guard will be knowing that the digital twin, not only can we know if it’s the digital twin, we’ll know it’s your digital twin and not someone else’s digital twin.

So we’ll be able to identify it as unique to you because of the sound spectrometer and the technology around this. We can track everything from your voice to your digital twin to someone else trying to create your digital twin.

Maureen: So. From a, let’s jump into banking as an example. The all of the situations where mom and dad get a call, Susie’s traveling and asking for money because she’s been abducted by aliens, whatever the scenario is, grandma gets the call that, that people are. [00:06:00] On purpose transferring money because of these deep fakes and frauds and the bank can’t necessarily protect if I have moved money of my own accord.

Brad: Yes. Well, with SGI technology, we are capable of knowing if the signal coming into the system is a, is a, uh, deep fake, or if it is actually a real human. So we understand we can gate that. So we, we would know, the agent would know immediately that the signal coming into the system is, this is not a real human being.

And why do we know that? Coming back to the fact that we can define what a vocal cord, what the DNA of a vocal cord is, we know exactly if it is a human being or if it is a, a fake voice coming in. And I think we could even go further to start tracking that if we were to set up some kind of a system to say, okay, now let’s, let’s set up the forensics on this and know, well, [00:07:00] let’s start, start finding out where is that audio signal coming from.

Is it coming from the phone number on file? Is it not coming from the phone number on file? There’s all kinds of forensics now that can be put in place based on the fact that we know that that is not a real human voice.

Maureen: So when I called my bank again, let’s stay with the banking example. They know my phone on file.

Brad: Yes.

Maureen: I, if someone has stolen my phone and calls in and says, I’m Maureen and I want to transfer X number of dollars to some other account with the deep fake detection, you would be able to tell if it was, uh, my digital twin doing that versus me

Brad: Absolutely. Not only could we tell if it was your digital twin, we could tell if somebody else created another type of audio signal based on not only if it was your digital twin, even if they had a recording of your [00:08:00] voice, we’d be able to know a. That it is not just, it’s not only not your di, it may be not your digital twin.

It’s somebody else trying to mimic you. That’s how sensitive the system is, and that’s how deeply we can get into the sound profile based on the technology that we built With the sound spectrometer, I.

Maureen: So from a risk management and a fraud detection perspective, this can be a game changer.

Brad: Absolutely. This is the next generation of how we handle and deal with, uh, fraud in the system. We are, we, we have now taken out that concern. That’s what we, that’s what, that’s why we’ve taken nine years with our head down in the trenches developing this technology, and we have just now come out of the black box in the last four months, five months.

Maureen: So what does SGI stand for?

Brad: Sound, genetics, Inc.

Maureen: And why is that name appropriate to describe what you do?

Brad: I think it’s appropriate because [00:09:00] we’re, we’re getting so deep into sound. So, uh, voice authentica, deep fake detection, voice authentication is really just the beginning. Of what we’re doing. This is kind of demonstrating the technological advancement that we’ve created here. We get, we can go. When you’re thinking about sound, uh, sound has, uh, a huge set of parameters to it.

Uh, your visual system has sort of one octave in terms of its ability to look at light.

Maureen: Mm-hmm.

Brad: sound. There are at least 10 octaves that go down into the sound profile. Our CTO, uh, an am JAR Cash, uh, developed this technology as he’s been studying bio acoustical engineering, not just bio acoustical, acoustical engineering.

Uh, since he was like 13 years old, he got his first college degree from Cambridge when he was 13. And so we were, we were fortunate enough to intersect with him nine years ago and, uh, start developing this technology. [00:10:00] So we know that a sound spec, a spectrometer is the ability to be able to separate spectra.

What we’ve done with, uh, with sound is we’ve now separated the spectra into, at least right now, 34 different frequency dimensions or frequency orders. So we’re going so deeply down into the sound profile. We actually believe we can get to the particle level.

Maureen: So what opportunities does that create businesses? We’ve talked about fraud detection. I assume from the level of specificity that you’re talking about. There are a

Brad: Well.

Maureen: different use cases.

Brad: There, there’s probably thousands of different use cases. Uh, one of the use cases that we’re looking at right now is we’re talking to a very large company about their hvac, uh, systems. [00:11:00]And so, uh, we’re talking to Batel, uh, out of Ohio. And, uh, Batel, we’ve talked to some of their engineers. They have, uh, servers full of HVAC audio data.

That has defines what the perfect operating, uh, sound is of a, uh, of an HVAC system. So our sensing capability can, we can now load those al design an algorithm, load those algorithms into what we’re calling now an acoustic net that we’re building. And so we have, we’re creating a product called the sound net, and these acoustic sensors will be about the size of this, you know, a little pod here and, uh,

Maureen: case for people

Brad: and you’re.

Maureen: looking.

Brad: Yeah, on your pod case, if you can’t see that. So a pretty small device. We’re actually now sourcing all the parts for this, but we actually now have a demo that’s available on an iPhone where if we loaded the HVAC sounds on it and it goes off of perfect, now we know. Hmm. It’s time to maybe have someone take a [00:12:00] look at the system or once we got the, uh, the original sound or the perfect sound from there.

Now we can take a look at all the sounds going down into the profile to where the system would actually need to be serviced. So that would be sort of the HVAC world that we could go into. We could go into car engines, we could go into plane engines, we could go into any kind of machine noises that need to be analyzed to know.

Let’s say if your BMW is all of a sudden there’s a fan that’s not working right or something else in the car is not working right, and I’m sure BMW’s defined and and has recordings of all their different. Audio, uh, uh, for their, uh, for their engines and all the parts and everything that’s running in a car.

Uh, drones, let’s say if your, your drone is not working. Any kind of machine noise, anything that has a sound. So that would be one way. Let’s say a building. We wanna, we wanna mic up an entire building with sound sensors. Now, let’s say we have a digital twin of the building, and instead of having cameras, now [00:13:00] you’re listening to everything in the building.

So now if the H HVAC system on the 13th floor, all of a sudden you’re having a problem up there and there’s a rattle, or there’s something that you don’t understand, now you can go and investigate it. And we call this digital audio imaging. Which gives us this, the ability to be able to see with our, with our ears, not with our eyes.

Maureen: So just thinking again of use cases, I just came back from a long trip and we had people whose planes were diverted for all kinds of reasons and we were all converging in Europe. If. The airline was using something like this in addition to tracking preventive maintenance. ’cause I’m sure they’re tracking all the parts and already do a lot of sensing, but not audio sensing. This gives me more precise data because the, part failure rate is based on [00:14:00] averages over time, not this specific. Thing. I, I actually worked with the, um, a missile organization years ago, and they were testing missiles before they went into Desert Storm. The missiles were created not in desert conditions, so parts failed because they weren’t designed for desert use. That’s a very short summary of a. Millions of hours of analysis, so take

Brad: so, so any kind of sound, any kind of sound analytics, uh, that, this would be in addition to, you know, augmenting anything. So we would be another layer. So we come in as another way to solve the problem. Another way to help. So any other kind of sensors like accelerometer data, vibrational data, any other data that might heat data, you know, any other kind of sensing devices that might be.

Associated with a [00:15:00] plane or with a car, or with a train, or anything else like that? Uh, yes. Uh, SGI technologies could be on board to give further data to give us what we’re calling audio intelligence. So the one thing about sound that’s different than any of the other, uh, uh, modalities out there, sound never stops.

Sound is always available. SGI, we’re not recording sound. We don’t record it. We record it to when we maybe create this algorithm for the sound, but we’re always listening. So what’s loaded on this device? What I’m, which I’m showing is the little, uh, the pod, uh, packet here. Uh, when it’s listening, it could be listening anywhere.

So let’s say for instance, on a border. Let’s say we create a sound net along the border, we can create a net, say 30 miles long by 40 miles wide. Anything that walks into that net. Connected. So these things can be connected together. These little sensors, we call ’em, the Sound net, can be connected together and they can talk to each [00:16:00] other.

We’re using Nette as a communication tool right now so that they all talk to each other. So as the say, we hear footsteps at a border. We hear people talking at a border. We hear trucks at a border. We hear people approaching a. 30 miles away, and we say, now we need to get to that pipeline, like let’s say a pipeline in Nigeria where there’s a lot of, uh, you know, fraud going on where they’re stealing oil.

What if it, what if the government could have a way to know in advance that someone is getting closer to the pipeline and that they, there’s a truck going there and they’re gonna steal the oil. We’re giving them a digital audio image. What’s around that area now, these little devices run on batteries and they can run for years at a time, and they will be very cost effective so that they can be deployed anywhere in the world.

They will not need any infrastructure whatsoever. Uh, they talk to each other. Maybe at the end of the net you’ll be, have a, have a wifi or a satellite connection. Uh, maybe at the end of either side of the [00:17:00] net. Um, now we can have intelligence on a border, very cost-effective intelligence on a border to be able to understand all the surface activity.

But we can also use accelerometer data. Which is vibrational data to know what’s going on underground. So if someone’s tunneling underground going around going, finding another way to get in SGI technologies can be built there as well. So there’s a huge, uh, variety of different tacks that we can take depending on how we wanna deploy this like.

Ports, uh, you know, we can say that’s a rat, that’s not a human in the back of a port, that’s a human in the back of a port. We need to go investigate. So we can also load these things on, on, on robots. We can give robots. Situational awareness. They have eyes, but they’re deaf, they can’t hear. Now we can add the ears to the robot.

So when the, a robot sees something, it can turn its head and look and see, and then we can tell it what it’s, what is it seeing [00:18:00] so we can create a awareness. We what the way Anmar designed the system, it replicates the human hearing system. So we, if you hear here, you hear anywhere behind you or around you, we can create the situational awareness for machines.

Maureen: So if I think about water systems, because water’s one of our right now most pressing resources and will continue to be so it could. Identify then where water was leaking potentially

Brad: Absolutely. Absolutely. And I’ll give you an example of that because we’re, um, I, I’m from Canada and we’re actually, uh, looking at a potential situation in Canada where they have, and, and Seattle, Canada, uh, Calgary, Alberta and and Seattle, Washington both use the same pipes. To for their water systems. And they’re about, they’re getting so old and now they’re getting pretty rickety, and so a lot of them are cracking [00:19:00] and breaking.

And so in Calgary, in the summertime, when these things break parts of the city are completely outta water. And so we could do preventative, uh, kind of work on this with our sound sensors and be able to know if it’s cracking. If, if, if there’s a leak where, wherever there might be a problem in the water system, yes, of course it can be used for any kind of in infrastructure, bridges, any, anything that needs attention, that requires sound.

Sound is the missing piece. Somehow people have not been really focused on sound. SGI is here to help solve that problem.

Maureen: So I’m just thinking of all of the places when you said bridges. Uh, presumably every com country has infrastructure issues with aging bridges, aging roads, aging water. My neighbor, my neighborhood has aging water pipes. over a hundred years old. Stuff breaks there, there [00:20:00]clay tiles.

Brad: Yes.

Maureen: it would be useful, it seems to be able to identify potential breaks before they break.

Brad: Absolutely. And, and that’s the whole point of SGI is that we’re always listening everything. We’re always listening to everything. That’s our job is to listen to any kind of issues. You know, if we, if you like, wanted to pivot to healthcare. Healthcare is a big one. Uh, during COVID, uh, we actually did a lot of research on, uh, uh, cough id.

We were looking at building it, and we were working with a team of doctors in Canada, uh, to build a, uh, cough ID product. And actually now we, we, we have already defined a collapse. So we’ve done this and for those of you who can see it. You can see my clap going behind me. So we’ve defined the clap. Now it would seem simple, but to define a clap and only a clap is a super complicated test engineering [00:21:00] process.

And so we now actually have defined the clap and we have a demo for the clap. Let’s go to, let’s go to a cough and look at coughs. We could define if you’re having, if you have a cold. Maybe you’ve got the pneumonia, maybe, maybe you’ve got some other malady with your, with your cough. So, uh, NIR is now working on the cough.

That’s sort of his next thing, uh, respiratory id. So if we needed to listen to somebody’s, uh, uh, you know, heart, uh, or see what their breathing is, like the app from the phone, just go to your neck. We’ll be able to get a reading as close as probably a doctor would be able to do with the stethoscope. Now, my caveat is that needs to have FDA approval, and we’ve looked into that and the process would be about two years, but healthcare is another area.

Anywhere there is sound. If it’s in the human body, if it’s in an animal, uh, we’ve been looking at cattle. Cattle. The way they chew their cut and the [00:22:00] sound of their cut is a, a very important, uh, health issue for cows. And, uh, actually, uh, companies do have sensors around cows. Merck has a special device, a neck device for cows that’s sensing heat and vibrational data to understand the health of a cow.

So SGI technologies fit into this whole array of, of anything to do with sound. And so we define ourselves as sound analytics and anything that has sound, we can create an algorithm for it and define it.

Maureen: And so we’ve talked about use cases as our listeners are mostly leaders. What do leaders need to be thinking about? I’ve heard deep fakes, I’ve heard cases for security, risk mitigation, fraud mitigation. [00:23:00] What would you advise for any of our listeners who are saying, that’s interesting, but what do I do?

Brad: Okay, so, uh, my call to action would be is if you’re in FinTech and you’re looking for a deepfake detector, give me a call. I have a deepfake detector ready for scale. If you’re looking for voice authentication. Married with a Deepfake detector, we have voice authentication ready to go to scale tomorrow.

Give me a call. Let’s, uh, let’s have a conversation and, uh, start the process of moving forward. How we’re, uh, going to, uh, to work with, with you. So we are currently looking for six to eight lighthouse brands to be able to do some testing inside of their systems so they can see, verify, and understand how our system’s working.

We already have third party validation on our, on our deepfake detector. And we now have, uh, uh, false positives of, uh, 99.99%, uh, not, I’m sorry, we, uh, false negatives of and, [00:24:00] and positives of 99.99%. The system is ready to go. It’s re ready for deployment. So we’re ready to have discussions, you know, and as Sam Altman has said, you know, in recent weeks how dangerous it is to have dfas out there and that there isn’t solutions out there.

SGI has a solution. We are ready to go. So anyone in FinTech, please give me a call. Uh, we’re able to, uh, start the process of getting our systems, uh, tested into your systems.

Maureen: And so, um, you’ve said it’s available on phone applications

Brad: It’s available on, sorry. Sorry, Maureen. It’s, it’s available anywhere we are. We have built this so that SGI technologies can be on Android devices. They can be on iOS devices, it can be on computers. It can be on servers, it can be anywhere. It can be integrated with any system in the world today.

Maureen: Okay, [00:25:00] I’m gonna do a timeout. Where do we wanna go from here? Because I don’t want this to sound like pitch.

Brad: Okay.

Maureen: how do we. Um, do we do something that’s a little less come by from me

Brad: Okay.

Maureen: little broader about, um, leaders start to think about, define their gaps,

Brad: Okay.

Maureen: kind of like what analysis would you have them do anyway?

Brad: Okay.

Okay. Okay. All right. Sorry. Okay, so go ahead.[00:26:00]

Maureen: Mm-hmm. Yeah.

Brad: Right.

Maureen: You know, as, as I think about my question, and Dan’s. talk about at this point, is this so important for you? I started this company

Brad: Okay.

Maureen: I am absolutely committed to health of our overall economy and, and finding solutions to fraud or something.

Brad: Yeah. Okay. Well, well, what I can speak to is how we actually got started on this whole project, which I think is quite interesting because, uh, the origin story was about the fact that Alex and I had developed a, uh, a national [00:27:00] education company. Uh, we were teaching children how to read. I, I actually brought a program down from Canada called Literacy Links.

That was the most sophisticated tool for teaching children how to read one-on-one.

Maureen: And.

Brad: this is where the passion really, okay. Sorry.

Okay. Okay. All right.

Maureen: so Dan, maybe we put this, maybe I start, let me ask the question, what was the origin story? And then we splice that to the beginning of the interview.

Brad: Okay.

Maureen: Yeah.

Brad: Okay.

Maureen: Brad, we’re gonna be talking about your technology, but let’s start with your origin story. Why did you do this, and how did you get here?

Brad: Okay. Well, um, Alex and I had, uh, had an education company where we were teaching children how to [00:28:00] read. We, we did this for 20 years. We had a national company. Uh, we were doing consulting work as well as working with school districts, uh, mainly in California, but it was a large scale operation. We had well over 500 employees and we were teaching kids how to read one-on-one, which is the best way to do this.

We brought down a program from Canada, uh, if 30 hours of one-on-one teaching children how to read, teaching them pH Eames, pH Eames are the core of being able to teach a child how to read. If you don’t understand pH Eames, you’ll never know how to read. So. Engineering. A friend of ours, uh, took a look at the tiles that we had using, uh, these phoning tiles, which represent the letters in the English language.

And so he saw these on a board and asked, have you ever, uh, thought of digitizing that? And we said, yes, we’ve tried many times. And, uh, with, uh, not being successful. So he decided to go take a look around and see who he could talk to. He talked to the [00:29:00] top speech recognition team at Google. And, uh, they, uh, he asked them about getting the code for phone, eams, and they told him that they don’t exist.

There is no code for phone, eams, no one’s ever done it. Uh, it was abandoned in the 1980s and, uh, uh, it’s never going to be done. That was a call to action for me for being very personally involved with teaching children how to read for 20 years. If we could find a way to understand phon, eams, and create a test for kids that we could automate the system because kids have to be done that test one on one.

It’s the only way to do it and to get the accurate results to know if they have proper phonemic awareness or not. So we went on a global search to find someone who was working on little teeny tiny sounds, and we found Anwar Amma Jarque in England, who was working on spray can nozzles for the British government.

And he was so good at that [00:30:00] he was able to identify the different manufacturers of spray can nozzles. When we introduced phons to Anmar, Anmar started working on a phony ID product. And we got down that product around, uh, 18 months into the project, realizing that we would need hundreds of thousands, maybe millions of samples to be able to make it accurate enough.

So we didn’t see the financial viability of that product out of the gate. So we pivoted to other products. And so we built a tune detector like Shazam and then Silicon Valley Bank. Pushed us over to say, build up voice authentication and deepfake detection. That’s what the world really needs. So our passion really started out of teaching children how to read and built into now into a technology that we believe will have a huge impact on fraud and, uh, and, and, and deepfake and problems with, uh, uh, fraud in the world today.[00:31:00]

Maureen: And so that dovetails pers perfectly with challenges and anticipated challenges executives are facing across the spectrum, thinking about the risks of DeepFakes ai and as AI becomes more sophisticated, the risk. Corporate risks continue to elevate and, and what you’ve created at SGI is a solution to a lot of these issues.

Brad: Absolutely. And I, I think, I think where, where we look at, the way we fit into the leadership is how do you want to deploy SGI technologies? Where do you wanna see it in your system? How do you want to, do you wanna put the deepfake detector? In what aspect do you want that to be the first thing that comes into the, when the audio signal [00:32:00] comes into your system?

Uh, should that be the first thing that happens to know if it’s a real. Audio signal a real human, or if it’s a deepfake signal that’s coming into the system, the deepfake gate as we call it, we feel that that would be a really important tool for FinTech companies, for, uh, anyone needing any kind of deepfake to make sure that it’s, it’s a real human that comes into the system.

From that point forward, we could integrate. With their authentication systems to be able to say, how do you want us to integrate? If you don’t want us to go like full scale with SGI technologies, we can overlay, and I give you an example of something like that. Say you’re using six digit codes, and I do this with Bank of America.

When I call in and they, I need to get access to my account or they need to verify me, they send me a six digit code. And they say, can you please read that six digit code? We could add a voice signature over top of that six digit code, which would be a very simple and uh, [00:33:00] straightforward test to say Yes.

That’s, that’s Mr. Diskin calling in so I can, I can see where we would be. We’re very flexible. SGI is capable of pivoting very quickly to whatever the needs of a company would be to be able to say, yes, this is where we want you to. Put the deepfake detector. This is where we want authentication. So we’re very flexible.

Our products have been developed as APIs and SDKs so that they can be readily deployed wherever the company feels that they have a breach or they have a need for, uh, a gate of some sort.

Maureen: So I’m thinking of two specific areas. If I am in risk and compliance, whether it’s cybersecurity or financial risk, or at a board level, I’m doing various kind of risk analysis, and that’s part of the and risk committee. It’s part of the CS CFO’s role. It’s part of the CISO’s role. So any one of [00:34:00] those might. Consider this as a solution to a part of their risk. Not only fraud, but any number of things like we’ve talked about with, identifying, sorry, beep, trying to find the right word. Um, identifying. In a waterline risk of cracking, identifying in a pipeline, risk of cracking or risk of theft, any number of. Reliability and breakage issues as I am, as I’m looking at my budgets for whatever the next year or quarter and, and matching dollars to risk factors, solution, this solution [00:35:00] is something that I should be seriously considering in my. Overall business analysis.

Brad: Yes, of course. I think, I think that where, where we come into that picture is wherever there is a need to detect an audio signal that is a threat. That is something that needs to be defined, uh, for a particular, like if it is a water break line or if it’s an HVAC system or if it’s a security issue where somebody’s walking in an area, in a building behind a building.

Uh, wherever. Wherever we don’t have video cameras, we could, we can always listen. We can always hear what’s going on. It’s that other layer that SGI has has defined and can define what the sounds are. So the other point about it is that this technology is not a super high cost item. So when you’re looking at SGI, we’re going down into sound, so we’re small.[00:36:00]

We’re really tiny and we’re very cost effective. So bringing this new, uh, layer into the system, we’re not looking at spending giant amounts of money to get it deployed. So it’s a very cost effective tool because we operate on very small. We can operate all the way to the edge, so we don’t need huge compute.

We, we are, we define small compute.

Maureen: So I’m thinking of various applications that use drones as an example, back to my sewer line to check the health of the sewer line. Civil engineering, looking at health of bridges, those kinds of things. I could drop an SGI sensor or set of sensors onto my drones and my robots leveraging the investment I’ve already made.

That’s looking, now it’s gonna look in here.

Brad: That’s correct. And, and, and what, what we do, Maureen, is we, we [00:37:00] bring, uh, we bring ears to machines. We, we make them here however you would wanna make them here. So we can, uh, we can integrate with drones, we can integrate with robots, we can integrate with your refrigerator. We can integrate with your washer, dryer, whatever, uh, you know, a manufacturer would be looking at for specific audio signals, sounds of the refrigerator, sounds of whatever machine SGI technology can be on board to be able to help understand the health of the machine, understand if there’s, uh, you know, you know, access to something.

Yes. Uh, my answer to all of your questions is, is yes, and I think. From, from our perspective, it’s, we can define that for any user and in terms of what they need, what is, what do they want to hear? And, and we will help them listen. We will help them understand what their, what is is happening in their environment with our technology.

Maureen: As I listen to you, I, I always go [00:38:00] to the immediate applications in my life so I could put a sensor detector and walk around my house and see what I need to repair or

Brad: Absolutely. Yeah. Abso absolutely. If, if your car was not running properly, uh, it, yes, we could, we could help define what the problem might maybe even define what the problem is. We will know what the health of the, of the automobile is for, from the manufacturer. ’cause I’m sure they have all of the audio data available for us to be able to create the algorithms.

And then from there, yes, we can find out if, if you’re having a difficulty with. Whatever machine it is in your house.

Maureen: So if I’m going to buy a new car or a previously owned car, I’m buying a previously owned house, which most of us do, I, this could be a layer on top of the home inspector that walks around and makes sure the refrigerator still has doors and is running, [00:39:00] but doesn’t necessarily tell me that it’s. Not going to be running next week.

Brad: That’s, that’s right, that’s right. The, the, we can go down that track as far as anyone wants to go.

Maureen: Mm-hmm. Uh, yeah, I’m just hearing multiple applications for only large enterprise, which I assume is where you’re targeting, but also for, if I can put it on my phone and add the. Uh, test my refrigerator, test my car, test my HVAC system in my house. Test my water lines. I, I live in a hundred year old house stuff doesn’t work sometimes. Um, I can, if, if I’m running a multi-family facility, how do I know if my apartment building needs an investment in HVAC system? [00:40:00] And how do I then build it into my budgeting process?

Brad: Absolutely. I, I mean there’s endless applications there in terms of how we can, uh, integrate. It’s just depending on what the needs are of a particular company and how they wanna utilize our sound technology. So that, from our perspective, uh, we’ve looked at many things and, uh, we, we kind of look at the core technology as being, uh, you know, the sounds spectrometer.

You know, again, the tool that you’re, if you can see it behind me, uh, that is what we’re using to define things. So, uh, we can create sound algorithms for, for anything. And, uh, honestly, I think the most difficult one that we’ve done so far was voice authentication and deepfake detection. Those two were super complex and took a long time to build, but I believe if we can do that, you know, I know if we, we have now accomplished that we can do anything.

And so we’re, we’re even looking at things like behavior, uh, you know, understanding behavior. So there are companies [00:41:00] out there that are looking at behavior models. So we can look at behavior models, uh, we can look at, uh, age models. I is this, we can add this to an age verification system to know if, if you’re, you know, we can tell if you’re a child.

We can tell if you’re an adult. So, uh, just depends on what you, what a customer wants to develop in terms of being able to, uh, you know, have a sound algorithm to define whatever it is they’re working on.

Maureen: Beautiful. I wanna make some joke about unhealthy aging, but I’ll, I’ll leave that for,

Brad: Well, well actually Anmar has, my mother is 96 years old and she sleeps a lot. And, uh, we wanna get some, some data. She has, uh, she has sleep apnea and that would be very useful data, uh, to listen to her breathing and to get that data and to understand, you know, what’s happening with her. Because, uh, her brain works fine, everything works fine, but she will stop breathing for, [00:42:00]you know, 90 seconds, two minutes, and then take a gasp of air.

And then she, you know, and then she’s fine. So this could be used in so many different ways. And, and the way I’ll explain SGI from its origin, we have what we call application paralysis and where we have so many different things we could go after.

Maureen: Mm-hmm.

Brad: And, uh, we are actually talking to the government as well about different, uh, potential ventures, which I can’t discuss that, that they’re interested in.

Maureen: Of course.

Brad: But, uh, uh, we, we we’re, we’re very versatile. So we kind of look at sound in is a open, kind of new world that if someone needs a sound algorithm, SGI can mint it for you.

Maureen: Uh, this is just fascinating and I’m gonna try to summarize ’cause I realize we’ve gone all over the place. But the idea that the risk of AI has [00:43:00] the risk profile for many of our companies and our individual lives, and yet also. opens up so many opportunities that just weren’t there before, and so what you’re creating with the sound side allows us to increase physical security, increase awareness of preventive maintenance in addition to things that are already being tracked, increase on the health side.

Thinking of. Listening for coughing and breathing and, and different things that we would be able to potentially monitor people with chronic diseases and, uh. Send them a text that they need to do, fill in the blank. X shock your mother when she goes, when she’s not breathing for a certain amount of time. [00:44:00] Um, but the, the applications seem broad. And from a leadership perspective, as we’re thinking of innovation, how do we leverage the latest technology to solve some of our business risks?

Brad: And that’s what SGI is here to do. So we are also ai, but our AI is built on SGI technology. So it’s, it’s, it’s our, it’s our AI based on the fact that now all of the data that we call it neoteric data, because this data has not been been collected before. So it’s new. And so collecting it and using it in a brand new way.

So let’s say it’s breathing data or how it’s a, uh, or it’s a cough data, you know, so there’s, uh, it’s uh, uh, any kind of, uh, sound, uh, uh, Anmar iss very interested in listening to all kinds of different things. He’s not, [00:45:00] you know, he’s not static and so the world of sound, in our view is as big as the world of sight and vision may be bigger.

Uh, he’s actually loaded into the system, the Sound Rover or the Rover on Mars. And, uh, we’ve, we’ve listened to things on, on the, on Mars that no one even knows are there. And so we, we’ve, we’ve, we can do, we can really see through our ears. That’s what we do.

Maureen: Which again for our listeners, opens up potential to solve problems that weren’t previously solvable. In this way, it’s much more cost effective and accessible.

Brad: Yes, yes, and integratable. Because the way the technology has been built is for integration into current technologies. XGI doesn’t want to replace anything. We just want to help make things better. We [00:46:00] wanna make your security system better. We wanna make your health system better. We wanna teach kids how to read better.

We wanna help the world move forward in its process, utilizing sound.

Maureen: Beautiful. Brad, how would people learn about SGII assume website, LinkedIn blogs. So give us that information.

Brad: Yes, uh, our, our, our, our website is sound genetics inc.com. And, uh, my email is, uh, is available on the site. You can reach out to me anytime I am, uh, you know, available. And, uh, look forward to setting up a, uh, exploratory session.

Maureen: Beautiful. Thank you very much. And to our listeners. Thank you for engaging and looking for opportunities to leverage the latest technologies to improve your businesses and improve [00:47:00] your impact as leaders. Please like us, share us, follow us, give us good reviews on whatever platform you you follow.

Brad: Thank you Reen for having me. I appreciate it.

Maureen: Thank you, Brad. It’s been a pleasure.

​