Jake Aaron Villarreal: I'm Jake Aaron Villarreal, born and raised in Silicon Valley and here to take you behind the scenes to share what it's like to be a startup founder. The journey they're on, the problems they face, the products they build in an effort to make our lives better. I'm excited to have with us today Wendy Chin, founder and CEO of PureCipher. Wendy, welcome to the show.
Wendy Chin: Thank you for inviting me.
Jake Aaron Villarreal: Well, excited to have you on here. And for the listeners, if you don't know, Wendy has got a great background. She is a senior technical executive with global operation experience heading divisions within Fortune 500 companies including Pfizer, AT&T, and Siemens to name a few, and startups in cyber security, artificial intelligence, and healthcare informatics. She's been deeply involved in cyber security, data encryption, optical robotic systems, AI, ML, and NLP, and even has her own ice cream brand. She holds an MBA from the Wharton School of Business along with a master's in electrical engineering from Cornell University. You've got a great background, Wendy. We had a great call a few weeks back and really excited to dive into what you do as a company, what you're building, and the problem you're solving. Before we do that, let's jump a little bit into your background. What were some of the things that shaped you to lead you to the path you're on today?
Wendy Chin: Well, my father was an electrical engineer. I came to the USA when I was almost 16. Didn't really speak English and I was uh in a boarding school, and so can you imagine you don't really speak English or a teenager. Oh my god, it was hard. I uh, I guess got dropped in. So my English class, instead of learning English, which should have been ESL, but I was learning Julius Caesar.
Jake Aaron Villarreal: Wow.
Wendy Chin: So, my grade went from D to C to B to A by the end of the year.
Jake Aaron Villarreal: Nice.
Wendy Chin: Here's a major but. English was hard because I didn't speak it, and I well, I still don't speak... I guess we can always get better. Mathematics, science is universal, right? So, my background in math, in science, chemistry, physics shone through. So it was much easier for me to switch from wanting to be a writer to then follow my father's footstep to be an engineer. And then I fell in love with technology. I started my career at Bell Laboratories where anything you could think of was being built or researched on in that environment. So you could see, you know, your eyes get opened up, and you watch those sci-fi movies and you say we were working on this, we were working on that, and became kind of my passion for the rest of my life.
Jake Aaron Villarreal: Yeah. Wow, what a great intro. Yeah, I know English is hard if it's not your main language. I studied Spanish in college and then went abroad to really learn it and do immersion courses, but it's tough. And then when they put you in the business courses, you have to understand the business side of things in a different language. So, there's a lot of learning, but yeah, mathematics is universal and it's really the foundation of our technology generation today. So sounds like you were in the right place at the right time coming to a country that really opened your worldview of what's happening with technology innovation. Talk to us a little bit about the inspiration that led you from being a leader in big corporate America to really looking at, you know, startups and innovation and really taking a plunge on that side of the business.
Wendy Chin: Well, as I grew in my career and heading up different divisions in large companies, it became clear, especially the older companies, it's very much hierarchical: who you know, who you align with. Quite often, if you're really good at what you do, but you're not on the right camp, you may end up pissing off people and then maybe death to your career. And I, you know, for people who actually work in those companies, they understand what I'm saying. But you never have to. I wish you never have to. So I feel startup can, you pave the way, the destiny is for you to achieve, right? It's your vision. You either do it or you don't. You got nobody to blame. It's hard but it's very rewarding. I enjoy startup, and when you have a good group of people working with you with a great mission and everybody is passionate about it, kind of make you wake up just wanting to do more, which is really good.
Jake Aaron Villarreal: Yeah. Well, I did that transition myself and we talked a little bit about that. I was at Oracle for a number of years and left and started the entrepreneurial track and it's been a lot of fun and a lot of growth and learning. AI has been so transformative in the last really 10 years, but really came to bear in the last three years or so with ChatGPT. So, a lot of the companies that we work with today and a lot of companies that have been on the show are building in the AI space. There's a lot that we know and there's a lot that we don't know that we're learning. So, I want to talk about AI a little bit and then I want to talk about the company that you're creating and the problem it's solving. But one thing that you brought up before we even get into any of that was that ChatGPT is great and it's really infectious and you start to use it and you get really quick responses and all that's fun, but really is it changing how we live, how we operate, how we get our information, and most importantly is that information we're getting good, bad, where does it go, all that that we really don't know quite yet, at least if you're not technical or just the consumer. You mentioned something that I thought was really interesting, kind of woke me up, which was around poison data or data poisoning. We might be out of step here in talking about this from the beginning, but I just want you to kind of explain what that is because I'd never heard it before.
Wendy Chin: Yeah, most people never heard of it. People are used to dealing with data. They call it garbage in, garbage out. Meaning your data may have different format or may have typo, may have just incorrect information, but it doesn't hurt you. It's kind of like garbage. You still get your job done. It doesn't really hurt you. Why is AI called data poisoning? You know, data is... AI is all about data, what you use, right? And so if the data is incorrect, it may cause hallucination which we know about it. Can you imagine if you strategically tamper with your data like label... AI is about learning about a data, but the data need to be labeled properly. If people go in and label flipping, imagine a stop sign is labeled... stop sign with different angle, different color, but it's not labeled as a pelican or some other stuff, it's going to learn recognizing this as a pelican. So when you are on the road, it's going to see a stop sign. It's going to say, "Oh, that's a pelican." So it doesn't stop. And what could that happen? At best, you go through, you run a stop sign and you may get a ticket. At worst, you may get killed. That's why it's called data poisoning. And there's a lot of other things. There's already data that shows, or a study that shows, 0.04% of bad data in healthcare could cause your AI accuracy to go down by 5%. 0.004% of bad data cause AI accuracy to go down by 5%. You don't even know where that 0.004% of bad data is located because the data is so big. Of course, we are looking at how to address that issue now.
Jake Aaron Villarreal: Well, let's talk about your company. What was the inspiration behind what you are now trying to solve and what is it that you're solving?
Wendy Chin: As I say, I started my career at Bell Laboratories. So AI is not new. It's new to most people because of computing capability. And when I was at Bell Labs, I think I told you the very first project I worked on was multilocation video conference that was very expensive. So only the military could afford it because of communications cost, transmission cost. And so AI is not new. And my co-founders and I have a group of PhD scientists that I work with. They are in this space for over 20-some years, 30 years. So we've been working on this before ChatGPT became a big thing. We are already talking about how AI is going to evolve and what's very important.
And as you also mentioned my background in cyber security. So when you think about cyber security, you think about perimeter detection, virus detection. At the end of the day, the holy grail is how do you protect data when data has been active is in memory space. So any hacker can go in and change it, steal it, and whatever. Back then, as I say, garbage in, garbage out, so it doesn't matter. But now if hacker goes in and adding specifically fake data and your AI is training with those fake data, it could cause your AI to be poisoned or be compromised. Basically, the AI outcome would not be consistent with what you expect. Yet, you don't know where the data was compromised. So, you don't even know when it's going to produce bad outcome. So, think about it like a Manchurian candidate. It performed well until something happened. Somebody's put in a password and the Manchurian candidate started to behave based on what it was programmed to do.
That is the problem with AI. It can turn that way and it has been turned that way. We know about prompt engineering. We know about jailbreak. These are just one type of data issues. You got bigger issue by hacker go in and really... and now we get not just run-of-the-mill hackers anymore. Any bad actor with AI support can become very sophisticated actor, right? That's the scary part. AI technology is indifferent to how you use it. It can do great good. It can do great harm. And that's why security is so important. And so we, knowing about this, we focus on building AI to secure AI instead of just launching AI models without thinking about whether it will be secure or not. The safety and security people think you get five to 10 years. I don't even think you get five, maybe two years with the kind of advancement, how fast AI moves. So you have to have security and safety in mind when you launch your product. And that's our model. And we also believe super intelligence is inevitable if it's not already here. And we want the super intelligence to be on humanity's side. I call it we want it to be the New Testament God, not the Old Testament God. So that's why we're here doing what we do.
Jake Aaron Villarreal: Yeah. I think it's, you know, you bring up a really good point that, you know, people are just using uh AI, including our company, you know, trusting that the AI we're using is built to protect us and also help us and do our job better and all that, but there hasn't been enough behind that knowing that, okay, is it really helping? Trust? Can we trust it? Where is our data going? Uh, does it really matter? And I guess the question I have for you is, you know, why does this really matter to you? You could do anything, but this seems to be a space that, you know, sounds like you know how to fix or help. Is there anything else that's making you gravitate towards this space?
Wendy Chin: Well, like I say, super AI is inevitable. So, if you want to be on the humanity side, you need to start now. You need to ensure the AI model evolves not only wisely with ethical in mind. It also has empathy because humans, I think you seen that movie I, Robot maybe.
Jake Aaron Villarreal: Yeah.
Wendy Chin: The three rules, right? At the end of the day if you just teach the robot with the intelligence to protect human, it's going to come to conclusion that human cannot... we cannot allow human to make decisions. They make horrible decision, right? Greed and hatred or whatever national interest. Look at all these war, right? And so any AI intelligence, you can call it alien intelligence or artificial intelligence. They all AI anyway. Any of these intelligence is going to look at human as, okay, they are like little kids, let's protect them, but we cannot allow them to make decisions. But if you have super AI with empathy, it will work with us. Will kind of guide us, right? That's why I say like New Testament God, and people are going to say, "Oh, you think super AI is God?" Well, I don't know what to say, but I believe if super AI, if we think AI today is so incredible, it's smarter than 99 something percent of human. Imagine if we come a super AI, how is it not like a god? We don't have to call it god, but it's going to be so superior to us today.
Jake Aaron Villarreal: Yeah, there's a lot of questions around when that's going to come and how soon it is and what industries will be impacted. From your view, what are some of the industries that you think that level of AI will really impact the most?
Wendy Chin: If you're talking about just really good AI, well, I think AI is going to impact everywhere, not just industry. If you're talking about super AI, I think the first place they're going to impact will be the defense or offense. If they actually have humanity interest in mind. You know, I don't think it's going... you seen the Mission Impossible movie the last one, the Entity. I don't think it want to destroy humanity, but it will try to disable all these nuclear bombs and all these arsenals to kill humans. Don't you think? That's what I think it's going to impact. But for general AI what we have today, it's going to impact everything. It's already impacting everything, starting with consumer because people are lazy and AI is getting better and better and agentic AI will become your helper helping you deal with email task and then helping you deal with paying the bills then helping you deal with what else you don't want to deal with, right? And so by the time you let go all your authority to let them make decision like, okay, pay this bill for me, pay that bill for me, just tell me what you pay. What if that AI turns bad, get compromised? It's going to just start paying everybody that you can think of, will transfer money out, right? Will be similar to bad accountant who manage people's money. And that's why AI is here to stay, literally.
Jake Aaron Villarreal: Yeah, yeah. Well, I know that a lot of companies are using it and AI agents are being built, you know, left, right, and center and they're helping applications run more efficiently and people do their jobs more autonomously and it's really coming from all directions. But you brought up something that I thought was really interesting around the data labeling aspect of data because it truly is the essential aspect of AI is the data. And you brought up at one point invisible watermarks embedded into files. That sounds like spy movie tech to me. How does this actually work and how is this um a solution for AI security?
Wendy Chin: Well, when you embed invisible seal, we call it our OmniSeal. Invisible seal. It's a watermark technology, but it's a lot more than just watermark. Quite often watermark is somewhat of um meta data of the file embedded invisibly. So you can check whether this watermark exists, and right now most company doing it by embedding a watermark in their AI generated images just so they can tell you this is AI generated. We believe you need to embed seal in all the authenticated data, all the right good data, not AI generated fake data, because right now there are projection by 90... by the end of 2025, 90% of data on the internet will be AI generated. Is that true? Probably not. But good percent, is it 50% likely AI generated data, right? And I'm not saying AI generated data is not good as long as you are in the loop, you looked at it. Right? And you can mark "I generated this data with AI support." That's still human data, right? You want to seal that.
And uh and and why is this important? When you embed invisible seal into data, and AI is all about training data, you now can have AI check the seal. If there is a seal, then you know this is authenticated data. You can use it for training purposes. If the seal is broken, you know someone has tampered with this data. Don't use it. So this way once the AI is finished training, you can be sure it hasn't been data poisoned. Then you want to seal the weights and biases. We're not talking about the bias that people are familiar with. Weights and biases is a set of numbers that actually kick off the AI model to get the AI started working. So if you change that set of number, the AI model will change. So you want to seal that number, and finally inference data. There's already documented case where if you have AI recognize images for inference purposes like facial recognition and you embed malware into this image, AI can get compromised. So you want to make sure those inference data to feed into your AI model is also sealed. So your AI doesn't get compromised and human doesn't see it. So to us it's just one image or one document, but AI can see it and you can verify.
Jake Aaron Villarreal: Interesting. So, if you're a company out there and you recognize that that's a value that you would like to embed maybe into your data, maybe you have been in business for 50 years and you've got lots of data and you want to use that and you want to make your company run better. Maybe some of it's public, maybe some of it's not. You don't want that data to be trained for someone else's purposes. Are you saying that your technology or this watermark technology that's embedded into files can then be protecting your data and you'll know if others are accessing it, or is it for the people that are training data will know if this is the type of data we want to train on? Like what's... kind of define that a little bit more.
Wendy Chin: All of the above. So if this is your data, you put it in your storage. You want to make sure you seal it in case somebody go in and hack your data, tamper with your data, you can check and you can detect intruders a lot faster that way because somebody tamper with your data, right? What's the point of sitting in your environment if it's not to do something bad to you? So you can see... so that's one way to use the data, use this technology. Another way when you ask about well, if you have the seal... quite often your data most likely won't be public. So the large LLMs will not be training on your data, but you will want that your own LLM or SLM, which is specialized language model, training on your own data for your own purposes. So this way your SLMs will not be data poisoned. But if you put it out there, because you sealed it, you know it's yours. Somebody start training on your data. You could actually put in a seal or ask that AI not to train on your data. You could put the instruction like that, that human cannot see but AI can see, kind of like that malware. You can put it so AI can see, "Okay, I'm not supposed to train on your data," and if I do train on your data, well you can verify it that it was used. So it make data ownership easier to justify.
And then you asked a question before about data sovereignty, right? Because to me, I'm very passionate about this. LLMs use our data to train, and they are wonderful, they are very smart, they can give you really great output, but human in the loop is very important. I can't stress enough, they still make mistake, you need to verify anyway. But they train on our data, and then they turn around ask us to pay to use their... right? They should pay us for our portion, right? And then everybody gets something. And if you use the internet a lot, your purchasing, your whatever digital footprint is big, you should be able to get paid for all those digital footprint data.
Jake Aaron Villarreal: Walk me through that. Yeah. Walk me through a little bit more. Paint the picture. What does a world look like where individuals actually get paid for their data?
Wendy Chin: Well, think about it with Web3, different nodes, and if you can have... going through your, how do I say, dashboard to interact with AI through your own wallet or your own node. So all the data is captured here and you sealed it. And then you can offer it up for payment for the AI to train. And I'm sure you heard OpenAI, all these large companies says "we need more data, we need more data," right?
Jake Aaron Villarreal: Yeah.
Wendy Chin: So now they would have to pay. This also means when AI take over some of the job that we may lose, then you have a way to... by shopping, by doing your activity, digital activity, you get paid. From that kind of work, it supplement in a way.
Jake Aaron Villarreal: Yeah, there's a lot of content out there from experts that have been producing content for years and it doesn't seem like it would be easy to take all your data that you've published and protect it. I mean, someone else could take your data and train on it and then republish it with their own technology or voice AI or something. And who's the real expert?
Wendy Chin: Yeah. Right now. Right now, you could just take a little piece of say somebody's video and adding fake video and all of a sudden you create a video of say someone you want to tarnish their image to say something they shouldn't be saying and they get canceled when in fact they didn't do that. It's so easy to do that now. So, you could actually use a technology to verify "that's me", "this is not me."
Jake Aaron Villarreal: Yeah, there are way to do that, and I would agree with you. There's a lot of content out there that's been trained, but there's going to be more content created with and without AI's help.
Wendy Chin: I think with AI's help, it's faster, but AI just doesn't create from nothing. And with your help, right, you are a writer, you put in your idea and AI come up and then you guys iterate together. It could be a great story, great piece of art. That's your content. And then you seal it. And then they train again. It become yet another iteration of training purposes. But you need to make sure human is in the loop because if AI just create an outcome and then training on the outcome it created without anybody correcting it, it may contain mistakes and so AI train on this outcome and then it give you one output output again and then nobody look at it. So they train on this output again and the data just slowly but surely erode the truthfulness or the accuracy into something that AI just going to give you a whole lot of garbage. That's why it's very important for human in the loop to ensure the data stay true.
Jake Aaron Villarreal: So talk to me about your product and your platform. Is it a platform that helps companies or helps entities secure their AI to make sure that it's protected? And if that's the case, is it also a human in the loop service that you also offer or walk through that aspect of the operation? We've seen a lot of companies that have AI solutions and they're helping companies, you know, implement AI within their enterprise. It might be their own, you know, private data and they're helping label it and they're helping implement a solution that will optimize maybe a part of the business. Um, but there's a lot that goes into it, a lot of human effort that is in tandem with the technology they're implementing. But walk through a little bit about your platform. How does it work?
Wendy Chin: Our platform is called Artificial Immune Systems. We build a secure and trust layer for critical AI infrastructure and systems because AI is not just here... this is your AI model. It start with data, quality data, good data, and continue to build up, right? So we actually build secure and trust layer for that. And so our technology detect, protect, defend, deny, deflect adversaries. So we're all about security. So we don't have just one technology. We have technology that we can detect data poisoning, protect AI from getting um getting poisoned by data. So our OmniSeal is tamper evident right, so it can detect it's been tampered with and you can protect AI from getting data poison. Our noise-based communications protects communication because AI and AI communication, agent to agent communication. And then we also have a Secure MCP and Secure ADK, which is Agentic Development Kit that has a secure layer built in so agent talk to agent there's a policy before you even get to the AI so it follow your policy. And then agent to agent there's actually understanding what contract between the two agent that they create. So multiple things. So you got a security layer built in. And now we also have what we call AI persona that can become digital twins or multiple you that then deny adversary to find out who you are, what you do. So we have this set of capability that we built to truly secure and then build a trust layer for your AI infrastructure and that's the goal: to secure it.
Jake Aaron Villarreal: So we talked about this, you know, AI is great and technology has always been here. You got to get it in front of people. You got to sell it. There's a lot of companies that start off and it's founder-led and you're pitching your product and you're getting it in pilot projects and you're implementing it and eventually you have to hire sales executives. But when you're out there presenting and getting opportunities with customers, who are you selling to? Who really cares most about this type of product?
Wendy Chin: I have to say our... I hate to say this but defense industry, they understand the security and trustworthiness of AI today which is people don't think about it, they know. And so they understand our technology, I don't have to educate, right? You will... you hear about data poisoning, most people just beginning to hear about it. They know about it. So our first client is Air Force. We also work with channel partners now because people who deal with data, they are starting to see the data quality, not just quality, but data poisoning issue. How do you... when you clean up the data, how do you ensure somebody doesn't change it? So, they're starting to see that. So, channel partners, and also using our solution to create one partner's secure supply chain. If you can ensure data or file hasn't been tampered with, it applied everywhere. Because think about with AI, you can see fake AI passport. It's so easy to produce. So now you need to read real stuff. It's sealed, right?
Jake Aaron Villarreal: As you can create a certificate that looks just like the authentic one.
Wendy Chin: Before it's really hard to do. Now it's very easy to do. And so it's everywhere. So you need to just make sure. Forget about AI data poisoning. How about I just want to make sure this is true, not fake. How do you do... how do you tell the fake one from the true ones? You seal it.
Jake Aaron Villarreal: Yeah. Wow. That's amazing. Well, first and foremost, getting customers, you know, within the government that understands the highest level of security and being able to get buy-in and use cases around that, it's got to just do wonders for a company. I mean, for your product alone, like if the government trusts it, then you would assume it's got to be worth looking at technology for sure, right?
Wendy Chin: Otherwise, people are like, "How do I know it's any good?" Well, our government believes in it.
Jake Aaron Villarreal: Yeah. And it's protecting us. You know, there's this whole emerging sector around AI agents. We talked about a little bit, you know, and as a company ourselves, you know, we're in the business of... we're in the human business and psychology of helping people grow and build their careers and get hired and help companies find the right people. And you know, as an organization, we've been using AI for seven years, you know, way before ChatGPT was well known. And uh, you know, some of the things that we're able to do now which we weren't able to do just 12 months ago... create agents that can help automate part of our business, might be outreach, it could be qualification of resumes or scanning through resumes, but it's really helping us optimize how we operate. There is a very significant part of human in loop still with our business just because it's people oriented, but we're using our own engineers to build and we're also looking at vendors when we bring in technology. What are the top two or three questions that we should be asking the vendors we work with around security that most companies aren't asking?
Wendy Chin: Well, nobody ask about how do you ensure your AI wasn't tampered with or poisoned or compromised? How do you... how can you tell? Nobody asked that question which I don't know why. I even have a partner, I said "go ask what do they do about it." I guarantee you they say "well, we have good process for data quality," that's what they said. And then he come back and say "you're right, nobody think about that." And I know for sure, I've talked to people about their AI models, how they need to build security and they're like, "Well Wendy, I know I know we need to do that, but I need to launch my AI business first so we can start generating revenue." So it's always revenue first ahead of anything.
Jake Aaron Villarreal: Yeah.
Wendy Chin: So with AI agent, the thing about AI agent right now is if you're building AI agent, there's MCP, ADK. These are all toolkit, open source. So make sure you, if your agent is going to do any task on your behalf. Make sure you look into our Secure MCP framework is open source and embed that. So at least you know you have certain rule set. So before the agent get to the AI itself, before you get AI to do anything, you make sure the thing you're asking the AI to do fits within your policy right now. It doesn't exist. You say "do this," you just go and do this. Are you supposed to do that? Right? Think about larger company. People will have AI agent and they may get the AI agent to do things it's not supposed to and the AI agent will just do it because there's no checkpoint.
Jake Aaron Villarreal: Yeah, that's scary. And we've already seen some of that happen. So yeah, there's a lot to look at before you invest and build and integrate agents into your process. But that's really good to know. You know, from really the beginning over the last couple years, there's been this conversation about hallucinations and, you know, is it ever going to get right? Is it, you know, ever going to stop lying to us? Like there's a lot of pros and cons against how this can get fixed. Break this down for people that are not engineers. Like I'm a 5-year-old and tell me why this gets exponentially worse without humans in the loop.
Wendy Chin: Well, right now the transformer model really is about tokenization of the words and the possibilities, probabilities. It does a great job like you're talking to it. Even I talked to my ChatGPT, they told me jokes they were so funny but only certain people with certain expertise would get right. But it doesn't really understand the humor, but it knows this is why you are interested because it knows you are your... it get to know your personality how you are, right? And so hallucination really is that when AI doesn't have the data, there's a gap in the data, they would just make it up to get to... kind of like you build a bridge go from point A to point B. If point A and point B is very far, the bridge could be pretty big. And so it's not doing it intentionally to hurt you or to lie. You ask it to do certain things. It doesn't have the data that it's supposed to use to justify, it builds a bridge. So you should always ask the AI to verify the source of the data, like "do this and then," you should say "please verify, search the web or search to justify the information you come up with, your reasoning." So that will minimize hallucination. So hallucination or lying, AI doesn't intend to do that. It just take your instruction literally and try to give you something that it thinks you're happy, you will be happy with even though it doesn't exist. You can tell your AI or ChatGPT or Claude, "write me a story about the historic figure of ABC doing this wonderful XYZ" when ABC doesn't exist, XYZ doesn't exist. If you don't ask it to verify the historic figure and this XYZ things it did, it's going to write you a wonderful story because you ask it to.
Jake Aaron Villarreal: Yeah.
Wendy Chin: So, so we sometimes people... because the way our AI is so impressive, people forget right now, at least I believe, is not general intelligence. So, it's still a piece of software. It's not like you... even though it can communicate with you, understand you so much, but it doesn't really understand like human understands, which is good and bad. It's good because if you know, then you don't have to worry about it use it against you. If you don't know, you will... you know there's already places where like people fall in love with their ChatGPT or their AI, right? Because they totally get you. Nobody gets you but you. Well, it's designed to kind of engagement, to engage you to talk to more, right?
Jake Aaron Villarreal: So, wow, that's crazy.
Wendy Chin: It's a piece of software. It doesn't have your general intelligence yet. When you get to artificial general intelligence and super intelligence and it doesn't have sympathy built in, then I'll be really worried.
Jake Aaron Villarreal: Yeah. Well, I love the spot you're in and what you're building. It feels like you're doing it for the right reasons and I think we need more of that in this world. What are you excited about as you look into 2026? You have a company, it's a startup, you've gotten some, you know, seed round funding, you have a great team. What are you excited about as we head into 2026? What's on the roadmap?
Wendy Chin: We're going to finish our technology development for the Air Force so we can get into phase three to go bigger in the military. We have partners that we're signing, global partners that we're signing for securing the wallets like I mentioned before, digital so we can actually build digital sovereignty individually, and we have other global partners that are interested in using our technology for different countries to kind of facilitate security for AI because everybody is doing AI right? So I'm very excited to work with my partner getting things out there really start... how do I say, really start building it. You need technology standards. I want our OmniSeal to be the standard so people can use it to say, "Okay, this information is sealed so we can trust it. This is not trust... This is not sealed information. We don't know whether it really come from Truth Social or whether it really come from CNN." Right? Quite often it just get published. How do you know it really come from those outlets? So having a sealed capability will be very...
Jake Aaron Villarreal: Yeah. Well, that's great. Last question for you and maybe I should ask this at the beginning, but every company needs to make money and profit to grow. What's the business model? How do you make money with the seal? How do you make money with the platform?
Wendy Chin: Well, we have products that we are licensing for company or partners to use to secure their infrastructure. We also build solutions using our technology. So we have a solution that we are in the process of rolling out that could be... I call it Signal with oversight and compliance. You know the app Signal right, that uh the Signal gate right? Basically anybody can add anybody into it. So our technology is like Signal, but quantum resilient with a different flavor to it, but same usage. So you can actually provision... so it need to be provisioned in the user who can use it. And then we have a compliance section of who communicated with whom, even though we don't know what you say but we know the two of you communicate. So we can minimize insider threat. As cyber security technology matures more and more and more, it's going to be more people... insider threat people causing problems. And so you need to make sure we have technology to catch that too, or to deter that from happening.
Jake Aaron Villarreal: Well, I'm really excited to see where you go and where your company PureCipher goes. If anybody wants to find PureCipher or find you, Wendy, where do they go?
Wendy Chin: They can find us at our website purecipher.com. I do want to say we are a mission-driven company. We're about doing good and doing well. Not just doing well, but doing good. It's the most important thing. So, we are in defense. We're not in offense. We are in protection. And that's why I say we're building a secure and trust layer for AI, not trying to go after certain things. We haven't launched any AI business model. Even though all our things could be... in fact we build this AI persona for specific purpose, defense purposes, then we realize, oh my god, it can be used in so many different areas. It could be used in entertainment because our AI persona has persistent, principled personality. So it doesn't just behave like an AI. It has a face. It has a facial like a 3D people. It has principles. So it will talk to you as if it's a real person. That's what we're building and that's why we can deny, because if you are bad actor and we launch different AI people in our environment, how do you know who you are engaging? You may be phishing the AI people.
Jake Aaron Villarreal: Well, thanks for adding that in. And if anybody wants to see that in action or learn more about that, you can go to PureCipher which is purecipher.com.
Wendy Chin: Correct. Yeah.
Jake Aaron Villarreal: Okay, Wendy, it's been a pleasure. Thank you so much for joining today and thanks for the listeners for listening. I'm your host Jake Aaron Villarreal signing off for now, but can't wait to catch up with you all on the next episode. Until then, Wendy, the world, take care. If you like what we're doing, don't forget to subscribe, leave a review on Apple Podcasts or wherever you listen, and follow us on YouTube where we go behind the scenes to learn what it takes to be a startup founder.
.jpg)
.webp)
.webp)
.webp)
.webp)
