What we talked about
Khurram Hussain: Khurram Hussain is the founder and CEO of IRVINEi and Mojo Solutions and Services, and he is building OVAL, an edge AI home hub that brings compute, computer vision, and voice AI directly onto the device. In this conversation, Khurram explains why the future shifts from cloud to edge, with edge enabling lower cost, faster response, and stronger privacy because sensitive data can stay local instead of being monetized by large platforms. He shares the personal story that sparked the journey: his autistic son ran out of the house and the cameras recorded it, but the system did not understand the situation or alert him in time. That experience pushed him to ask why cameras cannot deliver meaningful, real time intelligence, and led to the conclusion that cloud based approaches are too expensive for always on vision and advanced reasoning at scale.
Show notes
Khurram Hussain’s four-year-old autistic son ran out of the house toward the swimming pool, and the family did not notice for ten minutes. The Ring camera recorded everything but never triggered an alert. That gap, between what cameras capture and what they actually understand, became the founding problem behind OVAL, an edge AI hub designed to give homes and other environments the kind of real-time visual intelligence that cloud-based systems cannot deliver affordably or privately at scale.
What we covered
- Hussain frames OVAL the way IBM’s personal computer was framed in 1981: a new category that most people do not yet understand. Just as the IBM CEO at the time thought the world needed only about 100 computers, everything today runs through centralized cloud infrastructure, but Hussain believes that will shift as AI compute on the device becomes cheaper, faster, and more capable of protecting user data.
- The AI bodyguard concept goes well beyond motion detection. OVAL’s computer vision can distinguish whether a person at the door is brandishing a weapon, whether a grandparent who has fallen is moving or needs 911, and whether a child is approaching a swimming pool, all in real time, on device, without sending footage to a cloud server for analysis.
- Ring’s Super Bowl ad sparked a public controversy over the use of customer camera data to identify neighbors’ dogs without explicit consent. Hussain argues this is the inherent problem with any large cloud-dependent platform: the incentive to monetize data is structural, not incidental. Keeping compute and data on the edge removes that incentive by design.
- The software stack OVAL is built on did not exist before they created it. Hussain describes it as the equivalent of what iOS and Android provide for mobile apps, a foundation layer that lets developers build and deploy AI agents, models, and apps on edge hardware. He says they have already identified over 100,000 potential AI use cases the platform could support and are running a pilot with a small group of developers building on top of it.
- At the time of recording, OVAL had 350 customers pre-orders waiting for shipment, three signed distributors, and roughly 14 to 15 more distributors lined up. Hussain expected to begin shipping within 90 days, with an Amazon store also in the pipeline.
- The hardest part of building the product was not the technology itself but finding and holding together a team that is motivated by the vision rather than by near-term financial outcomes, a challenge he says Meta and Google would describe as one of their core ongoing difficulties, even with far greater resources.
About Khurram
Khurram Hussain is the founder and CEO of IRVINEi and Mojo Solutions and Services, where he has spent more than a decade building companies around emerging technology with a focus on AI and practical edge computing. OVAL, their flagship product, is available at hellooval.com.
- LinkedIn: https://www.linkedin.com/in/khurram-hussain-1673b4383
- Website: https://irvinei.com
Episode 122 of the PreVetted Podcast.
Full transcript
Federico Ramallo (00:00) Welcome back to the pre-vetted podcast where we spotlight extraordinary people and remarkable talent reshaping our world. Today we are joined by Kouram Hussain, he’s founder and CEO of Irvine Eye and Mojo Solutions and Services. Kouram has spent more than a decade building companies around emerging tech with a focus on AI, product, and building practical solutions.
At Irving Eye, he’s building oval and edge AI home hub designed to combine an AI bodyguard, assistant, and smart home connectivity, keeping AI private and running locally on device. Graham, welcome to the show.
Khurram Hussain (00:43) Thank you Federico, thank you for inviting me.
Federico Ramallo (00:46) So can you tell us about oval?
Khurram Hussain (00:50) So Oval is our product, right? And basically, as a company, are Edge AI company. And there has been evolution on the process of what we thought we are and what we have become. So we are an Edge AI company where we provide the compute on Edge. The problem with the, there are a lot of advantages of Edge versus Cloud, and they will grow more and more in future.
But before I go there, the problem with the edge computing, why it’s needed and people are realizing it very fast and they will keep realizing it. And we want to be in the forefront of that market. And what we have solved is we have developed the whole software stack besides the hardware stack for edge, right?
And when we say people think it of Oval or Erwini as a smart home and security company, smart home and security is just one application of our platform. We had to have an application of our platform because we say we have created a new category called AI Hub, right? So AI Hub is like a personal computer of 1981 or at least 1980, right? Simply because people didn’t realize what it is. Everybody was in the cloud with the mainframe computers working in the
and even the schools had access to it. But the people sought IBM CEO at that time said we all needs like a hundred computers or something, right? That’s what his idea of completing was that time. so right now everything is going through the very similar phase where everything is in the cloud and the applications developed are in the cloud because of their nature and size. That will shift faster than we think that the computer will be required on the edge and the people will be
understand more technology and more applications of AI will be developed which are more private and cheaper and faster just like a personal computer.
That’s where we come in as a J.I. and oval is our first application. Yes, it does not have a keyboard and mouse and like because it’s our AI will be more have a feature interface of voice and it will able to see us and we’re able to talk to it. And that’s what where all comes in because all does have a communication voice AI working on it. It has a camera and it connects with any camera. So it has a computer vision on it and compute on it.
in the future as we grow our future versions will have a local language models on it as well so even for the language models we won’t have to reach the cloud and it might sound very difficult right now but sooner than you think it’s a lot faster
Federico Ramallo (03:24) Right, right. Yeah, because the LLMs are developing so quickly. There are so many options now that that’s going to be more feasible in the future.
Khurram Hussain (03:34) Yeah, a lot of them are already feasible. It’s a matter of how you, if your software is capable of doing that or not. So.
Federico Ramallo (03:44) So why did you choose H-AI instead of doing everything in the cloud?
Khurram Hussain (03:50) The biggest differential is the cost, right? So here’s my personal story that my son was four years old, he’s autistic, he ran out of the home and I could not find him. And it took us 10 minutes to realize he’s gone and we started checking all the cameras and it was in front of the cameras, he’s running out of the home, towards the swimming pool. And we do not know and okay, we found him later. But he said, why the hell the camera didn’t give me alarm because the damn ring keep telling me there’s a guy or there’s a motion, right?
I want my own sense. So, and then we started going down the rabbit hole of why, and that why came on discovery with like the compute is very expensive, cloud is different, we don’t have any software architecture, we don’t have any hardware architecture, and doing that in cloud is prohibitively expensive.
and anybody who’s tried to do it in the cloud won’t be to do it for forcibly future and it’s getting only more difficult with the
with a shortage of data center availability and construction, right? So then we said, okay, this can be solved on Edge. We can do a lot of Edge. We tested and benchmarked a lot of GPUs and we found like, okay, what we can do and what we cannot do in the Edge and where the future is going. So based on that, we came up with this whole thing, doing it on the Edge. Because what we are doing in the Edge, our competition will charge you two, 300 bucks. They’re copying us. Ring is copying us day and night. If you had the CES, they’re using our features
names since our patent came out last year in May. They came out with a lot of features which are like they try to copy us left and right but they don’t have the hardware capability to do that.
And in our core, we’re not a smart home company. future is like our technology is deployed in senior care living to offload the nurse’s workload and the hospitals to inform doctors and nurses on the nurse stations properly because the applications of a technology are far beyond a smart home.
Federico Ramallo (05:41) Right, so Smart Home is one of the applications, but it could be using so many others.
Khurram Hussain (05:46) Yes, that’s what the core of the technology is. And with our open platform where the developers can come and develop a lot of AI agents and AI apps on our platform. And so the possibilities are huge once we look at it.
Federico Ramallo (05:59) And what does traction look like for you right now?
Khurram Hussain (06:02) We are going in production. So we are in the last phase of production to smooth out all the hardware things about the Edge, which we have developed in the last two and a years and put it in the production. So pretty much we are close to going into production. I keep saying 90 days, but this time it will be 90 days.
Federico Ramallo (06:27) So in 90 days, you’re launching to the audience, to the public.
Khurram Hussain (06:31) 90 days we’ll have shipments start. Yeah. Shipments. Because we have a 350 customers already waiting. So we have to start shipping to them. We have three distributors who wants to our product. Who we have signed alloys with. And we have about 14, 15 distributors who are lined up plus we have to turn on up Amazon and other stores as well.
Federico Ramallo (06:36) Very nice.
Right.
What is one thing that most people misunderstand about AI in smart homes?
Khurram Hussain (07:04) So, there are lot of misunderstandings about AI in general.
how AI will evolve this is very amazing and how AI will be used. So people are afraid of AI, rightly so once you have it in the cloud. And I think you saw the Super Bowl ad in USA, it is a big controversy with ring that they’re using other people data to find other people dogs without their consent. And they say they have a consent. Then today it’s a dog, tomorrow it’s spying for ice. And the day after tomorrow is like 1984.
or George Orwell everywhere, right? So that’s what the, because Ring is inherently is a big corporate company has been using the for data monetization at a very huge scale already, including working with other companies, agencies. So the problem are inherent with any big company which has to monetize the data.
So if the data is on the edge and you can secure a lock your data, that’s a different ball game. So that’s where we come in.
Federico Ramallo (08:04) Right, right, interesting. So you also talk about AI bodyguards. What type of functionality are you thinking ⁓ for the future?
Khurram Hussain (08:14) So in our smart home
space, once you give a brain to with the computer vision, you can have somebody who can look after you and tell you.
or what is going on. So when we say bodyguard, it means somebody’s trying to break into your home. There’s a guy with a gun on your door. This guy’s a policeman or this guy somebody you need to be worried about if the guy’s brandishing the gun. We do all that analysis in real time. Your grandparent fall down inside the home and if they’re not moving or you should call 911 or they’re fine. It looks like walking or child is going near a swimming pool.
right? Or somebody’s eavesdropping or somebody you put on your unwanted list and is stalking. So it’s sort of a bodyguard for your life where you don’t have to worry about your personal safety and security and somebody tells you in time.
Federico Ramallo (09:08) Interesting.
Khurram Hussain (09:09) But because you have to realize that our device connects with all the cameras around it, right? So it becomes a brain and watch for you because all these cameras are dumb cameras. We make them smarter.
Federico Ramallo (09:22) Right. They just recorded the image, don’t analyze it. Don’t understand what is going on. Right. Yeah. I’ve seen people using, the cameras with zone. basically if you have a wall, right, you can detect moving on, you know, outside of the whatever house, property and then inside of the property. then they can detect when you kind of somebody walk over the wall.
Khurram Hussain (09:28) No, no, no.
Federico Ramallo (09:49) Right. But the…
Khurram Hussain (09:50) That’s what we can do as well. Somebody jumped across the wall,
or we detected somebody’s trying to jump across the window. We inform you that somebody’s trying to jump.
Federico Ramallo (10:00) I mean, they do it based on pixels movement, right? Which is a very crude way to do it, right? But with your vision engine, then you can do much more advanced behavior detection, right?
Khurram Hussain (10:16) Yes, lot of it is like boundary reduction. You create a boundary on the pixel and anything moving across that boundary. We do totally differently. We have our AI for that.
Federico Ramallo (10:26) Right. And would developers be able to build on top of your platform?
Khurram Hussain (10:31) That is our plan that we believe that AI is not just one.
One big models cannot serve the whole humanity. It’s a lot of AI models and AI agents working together. And that’s why we, how we see that democratization of AI will take place beyond one, five big companies running big LLM models. So that’s how we will open up our, we have already testing which we are in pilot with few developers on.
developing AI apps and AI models, AI agents basically, and AI models working with them on our platform besides IoT devices as well.
Federico Ramallo (11:06) Right.
Khurram Hussain (11:07) So in future, we’ll have our whole AI app store where people can develop different apps and share with each other.
Federico Ramallo (11:08) Ahem.
Interesting. And what type of AI apps would you like to be built on top of your platform?
Khurram Hussain (11:24) So there can be like, it’s the same question if you ask somebody in not somebody, it’s 2007, you’ve asked Steve Jobs what kind of apps have we developed.
You can give some idea good ideas, but you can never foresee the whole human ingenuity which comes with it. So there will be apps which will be for safety or factory floor or nursing care and hospital and parking lot monitoring and factory floor. And there will be much more beyond which I can imagine today because a lot of those apps which we see like mobile apps, they needed the domain expertise for fitness app or communication apps or
organization apps which we use in today’s life on our own.
phone. did not. Those were developed by some of the people who had the expertise in this domain. It was difficult for anybody else to develop and deliver those quality apps to the people or the consumer or the users without knowing all those problems which the consumer needs to go through. So it’s a similar situation here because each problem is different in nature and people with the right amount of expertise and the data will be able to develop that use cases more better than one big company or
one small company doing all trying to do all of it.
Federico Ramallo (12:37) Right, right. It’s the, can vision some apps, the world can provide, can exceed our wildest expectations, right? When the creativity starts.
Khurram Hussain (12:47) Yeah, yeah, we did write a white paper and we
found in that white paper there are 100,000 plus AI use cases which can be developed using our platform. And this is just, this is just 100,000 which we counted, there’s like millions out there. So we just not know exactly, but we know the potential is huge.
Federico Ramallo (12:58) wow.
Right. Right. And what is your market vision beyond smartphones?
Khurram Hussain (13:14) You mean smart homes?
Federico Ramallo (13:17) Smart homes, yes.
Khurram Hussain (13:19) So smart homes will be a natural, our platform is really to tackle any kind of models for any industry today, right? Our voice control can be upgraded according to the needs. Our agentic AI system will be or somehow is capable of.
understanding the context based on the data and agents working with it and able to produce the results right so so technically we are ready on on all the on ai level the graphic inter our GUI the graphic user interface we will have to update depending on the industry if it’s required but without doing it if you are able to communicate with it
talk to it and develop it. It’s already the platform is already ready.
Federico Ramallo (14:07) Interesting.
Khurram Hussain (14:08) That’s beauty of AI, AI, you don’t need to design a GUI a lot. You can talk to it and if you can understand it, done. You don’t have to design another interface to adapt to the new environment, new industry or new use case.
Federico Ramallo (14:22) Right, can just use the chat interface and then everything else can be done through the conversation. Yeah.
Khurram Hussain (14:32) Yep.
Federico Ramallo (14:33) So if you could give a piece of advice to a founder building AI products today, what would it be?
Khurram Hussain (14:41) Well, I will not qualify myself yet to give a good advice to any founder. I’m in the process myself, right? But few simple things that you see that you have to go through it. should have a passion because it’s very painful.
time-consuming so but one thing if you are in it and you don’t feel that pain and if the pain comes and after a few moments you say fine we’ll go get through it every single day then you are the right person and do it don’t wait
Federico Ramallo (15:11) Right, right, amazing. So we talk a little bit about Oval. Can we talk a little bit about Irvin Eye and what problem, scope of problems are you trying to solve with that company?
Khurram Hussain (15:26) Yeah, so this is a evolution journey. So when I was a company started and we said, okay, this is the problem we are solving. OL was our product, hardware product.
So once we did the rebranding of everything we became, our new website is helloovl.com. Still there, when I dot com you write, you go to helloovl.com now. So we rebranded our company around the product rather than the product around the company. There are two methodologies of branding. you go to the marketing folks, that’s what they talk about, right? That’s a foundation of branding. So we rebranded the company around the
which is Oval and Irvine Eye is the company which is a HCI company in its core. Maybe we have different brands than Oval in future. Right now we are focusing on one thing which is Oval and our website is hellooval.com and foreseeable future that will be the case but maybe in different industries if you want to enter you need to have different versions of the product then we’ll have maybe different brands coming in that case.
I don’t know yet, to be honest, but we want to enter different industries as soon as we can to prove, to do the proof of concept.
Federico Ramallo (16:43) Right, right. So it’s…
Khurram Hussain (16:44) But the core platform
will always be old because that is where compute and the software has already been developed, all the whole software stack for Edge. So we need to understand one thing that if you want to develop a mobile app or Android or iOS, or you want to develop a web page.
you have a software stack which you follow to do those things, right? There was no software stack for Edge available before us. And we are the company which we have a software stack to allow, run the Edge AI and compute and AI agents on one platform, right? On the Edge. And that’s what differentiates us from…
We have a lot of companies. If we make that software stack public, that’s another way we can be very quickly grow as well.
Federico Ramallo (17:33) Right, right. Interesting, interesting. Yeah, it’s part of the discovery journey, right? That you build one brand, but then you pivot, and then you start building a product, but now you have both. You’re taking care of both and working on launching the product, but also having the company to support it. Yeah, I think that’s very interesting.
Khurram Hussain (17:57) Yeah.
Federico Ramallo (17:58) So what was the hardest part of turning this from idea into product?
Khurram Hussain (18:05) I think so whenever you go on a journey, is a lot of R &D, we have a lot of technological patents on our technology as well. So going through the R &D and finding the right team and the right folks who align with the vision and who love it.
takes time to grow then especially for a small startup because in the startup, always anywhere in the world is the focus. The focus and alignment is very difficult things to continuously prototype, do the R &D and change the prototype and to reach a stage and you have to do it so many times to do one thing because there’s so many features working, every feature has a separate prototyping being done.
It’s not an easy process. You need a good, very good team and focused team to do that. To establish that team and make sure that team is not only motivated but also aligned with the vision in the long term and they are in it for the love of it more than money and success which comes with it as a byproduct. That is very difficult. It’s not easy. People think it’s easy to do, but no.
it’s not easy to do. Even for companies like Meta and Google, that is the biggest challenge to date. If you ask them at the core of it. So that particular team, the good thing is we have it and I’m proud of it.
Federico Ramallo (19:17) Right, right. It’s super hard to build products. And one of the things I find hardest, the most, is to figure out what not to build. To find that right path and what are the things to avoid. I think that that’s hard, right? And with more resources…
you can afford to make more mistakes, right? But you’re walking a fine line of limited resources, trying to launch on a particular time, delivering on the promises you make to the potential clients. It’s a balance that it’s hard to do, yeah.
Yeah. So what are the plans do you have for oval and earring eye after the lunch?
Khurram Hussain (20:10) So we have a lot of plans, a lot of things we can discuss, a lot of things we cannot discuss, right? So we have come up with a whole roadmap beyond launch, which we have already planning somehow, putting it’s only from the drawing map to the pipe.
Federico Ramallo (20:17) Right.
Khurram Hussain (20:27) We are working on putting it there. We have plans to… It’s also depends on the evolution of the AI, how fast it can be and how fast mass adaptability comes in with it.
So there are a few very different plans in different directions. So it’s so much you can do right now in so many different ways, which can go right and wrong both ways. So in AI world right now that you…
that you have something which I have right now in in our pipe after six months, will be that we’ll be working on or there’ll be something which will be different. But whatever we do, it will be based on the foundation of what we have already developed with Edge AI solution. So it can be a different application, but it will be based on the foundation of Edge AI hardware which we have developed.
Federico Ramallo (21:21) Right,
right. Amazing, amazing. I’m looking forward to see the product, you know, and play with it and see how, you know, how it performs. Happy to give you any feedback when I try it out, yeah. ⁓
Khurram Hussain (21:35) I’ll definitely
sending out. I have one of the units which we received from China in the testing next room being tested right now. So we have more units coming in so there’s a finalization phase we are going through. So we can start the production and start shipping as soon as possible.
Federico Ramallo (21:41) Right, nice.
Amazing, amazing.
Khurram Hussain (21:53) And one thing I don’t want to be remembered our company as a smart home insecurity. It’s basically a compute provider which provide the edge computing stack whole stack hardware to software right now.
Once you look at it from one industrial application like home and security to edge computing stack to develop run hardware and software, it’s totally different two different directions. And I believe the edge compute hardware on the edge will be a lot of utility and utilizations is available.
And that’s where we will have something which we have some vision for, but market and movement around the market, different consumer will decide, we’ll be able to make that vision more clear. So.
Federico Ramallo (22:44) Right, right. Yeah, I think that people associated with one particular category because they tried to associate it with something that they’re familiar with, but…
Khurram Hussain (22:55) Yeah, because
here’s the thing. So if we say we are defining a new category of the product, right, which is AI Hub, nobody understands what AI Hub is today, but maybe six months to two years down the road, three years down the road, people will have it because once you say, okay, why I can’t do this with my this device or my computer or my ring or Google Hub, you say, no, it’s not an AI Hub.
And that will change the conversation or understanding of the people what AI Hub means. So till then, are in a phase where we want to catch up into that market faster. there will be, AI Hub will be a thing for all the businesses, homes, everyone really soon.
Federico Ramallo (23:30) Right.
Right. It’s the challenge of creating a new category, as you say, because people don’t know where to position the product because they don’t know what they… Right.
Khurram Hussain (23:51) Because our
minds things in the boxes. We want to fit everything in a box and if it doesn’t fit in the box it goes out of the box. So that’s the challenge.
Federico Ramallo (24:00) Right, right. But the payoff is that, you know, on one side you have to share the vision with the audience, with the market and convey that, educate them on that vision. And then the payoff is that once they get it, then people are going to be much more interested in the product, right? But it’s going through that initial
know, friction until they get it, right?
Khurram Hussain (24:23) Yeah, that’s
what our strategy was to have one category instead of finding out and then move. Then the challenge will be to detach from that category and being a mothership, not the subship. So we will go through that. It’s not a big challenge. If you offer our audience the right value, they will understand what we are offering. As soon as they can see the value.
Federico Ramallo (24:51) Amazing, Quran, I truly appreciate you being here today. Any final remarks before we wrap it up?
Khurram Hussain (25:00) Yeah, wrap it up. Yeah, AI is a fast evolution and we want to be in the forefront of it in different ways. So democratization of AI will define the future of humanity in so many ways that it empowers individuals without the control of big corporations and that’s where the world will change and it’s already changing.
applications of this technology on LJIS from home to the factory floors to the battlefields to everywhere. and we say the next thing is go to hello world.com and check us out.
Federico Ramallo (25:40) Great, great, we put the link to the website in the description below so people can go and check it out as well. Great, thank you.
Khurram Hussain (25:48) Thank you very much. Thanks very much.
Federica, nice talking to you man.