Ask a Different Question

AI agents don’t just answer questions, they take actions. That’s the shift that keeps a lot of security leaders up at night, because when software can move fast enough to outrun review, trust becomes the real control plane. We sit down with Jaz Lin, founder and CEO of Skyrelis, to talk about what it looks like to build AI agent security when speed is the default and quality is too often optional.

Jaz’s path is the kind that only makes sense in the rearview mirror: aspiring journalist and chief editor, then network engineer, CCIE, security leader across major vendors, and finally someone who has lived on both sides of the table. We dig into why journalism and cybersecurity share the same muscle: working the facts, challenging assumptions, separating false positives from true positives, and staying honest about what you can’t prove yet. That outsider perspective shows up again and again as we talk about identity, belonging, and how to keep your own voice in an AI era that rewards the most predictable pattern.

Then we get concrete about AI governance and AI runtime security. Engineering teams can ship agentic workflows at a pace security teams can’t always see, creating a growing visibility gap. Jaz explains why prompts and instructions aren’t enough, what “black and white” runtime controls could look like, and why she believes policy should be decoupled from applications so it can be enforced consistently across cloud environments and SaaS products.

If you care about AI security, agent visibility, policy enforcement, and keeping humans accountable for decisions, this one will sharpen your thinking. Subscribe, share this with a security friend, and leave a review so more builders and buyers can find the conversation.

This episode is brought to you by Onyxia Cyber — the Operational Cyber Resilience Platform that turns security data and asset intelligence into action. Onyxia helps security teams spot risk early, sharpen their strategy, and tie day-to-day operations to measurable business outcomes. Trusted by Fortune 500 leaders. Book a demo and mention Voices of the Vigilant to get your first month free → https://www.onyxia.io/?utm_source=podcast&utm_campaign=vigilant_violet

Transcript

Copy Transcript Copied

Jess Vachon: 00:33

Hey, welcome back to the show. This is a show where we talk to people behind the practice, the whole person, not just the resume, because the most interesting thing about anyone is security is almost never the part that fits on a business card. My guest today is one of those careers that only makes sense in the rearview mirror. She's the founder and CEO of Skyrelis, an AI security company building runtime security and control for AI agents. She's a network engineer by training, earned her CCIE, spent years at places like Cisco, VMware, and Rubrik, and she sat on both sides of the table. The builder shipping the product and the buyer in the CISO organization deciding whether to trust it. But here's the thing that got me. Before any of that, she was going to be a journalist. And I don't think she ever fully abandoned that. Jaz Lin, welcome to Voices of the Vigilant.

Jaz Lin: 01:24

Thank you. Thank you for having me here.

Jess Vachon: 01:27

So, let's start there because I think most people who end up running a security company did not spend high school and university writing for magazines and newspapers. You did. You became a chief editor. Take me back. What pulled you toward that world in the first place?

Jaz Lin: 01:43

Yeah, yeah. Well, first of all, I'm very impressed by the intro, and you go into a lot more detail than what I share with you. Looking back; I don't even know how I got here. It's not like my whole life so intentional. A little bit of luck, a little bit unexpected turn, and a little bit of what I want to do at the moment. So, yes, if you asked me earlier like in my teenagers, like in my early 20s, what I want to do, I probably would tell you that I would become a journalist. In high school and university, I spent a lot of time writing newspapers, writing for building the magazines, and eventually I became the chief editor. It's actually very fun. I love it. And what really fascinates me is not really writing itself, the story behind the story. Like if it's exactly the same events and two people writing it, you probably get very two different stories. So that is a part that really fascinated me. And throughout the weekend, you asked me to feel out of the form and like ask myself two questions. What about me? What about my background that I look at differently now than before? Second question is what I did, I truly enjoyed. So, there's another part of my background. I do think that I look at I look at it very differently now than when I was younger. It's an outsider part of me. I am my father is veteran and came from China. I grew up in Taiwan, and my mama is Pacific Islanders in Taiwan. So, we're not a typical Taiwanese family. And then later I moved to the United States as a queer. No matter where I go, I always feel like a little outsider in a group that I belong to. So, what does that really mean? I ask myself when I'm walking, walking my dog, you know, preparing the phone for you. I think the couple of things I keep reminding myself, like I always want to make sure myself like the most common way we say things, not really the only way we say things, oftentimes like behind the closed door when I spend time with myself, I always want to remind myself to keep my own point of view, my own perspective. I actually find that like that's especially important, especially today in the AI's era. Like AI is just like so good at finding patterns, I just don't want to become that common patterns. So that's one thing I look at it very differently. I very, I mean, appreciate that background, the outside background. And in somehow, I become an engineer, not a journalist. But one thing, another question I ask myself is that what I what I do, I truly enjoy, make me feel excited, make me feel happy.

Jaz Lin: 04:42

I think like growing up, I always like to fix things. Like growing up in Taiwan, Taiwan is very, very humid. We have a typhoon season; we have a rainy season. The phone line out in our house, the phone lines always get very, very wet. They got a lot of noise. I remember going out to follow my father up and down the house, trying to figure out which part, which session of the phone line is get got too wet, we need to replace. I remember that feeling, just like I feel accomplishing that feel proud, I fixed things, then now everybody can use a phone. I almost feel that that part of me, that's why I enjoy building product, solving problems. So that until today, that's still so true. Like every time when I solve a problem, I feel very proud. I feel very happy. Maybe you know, solve people's problem. Yeah, I think to be fair, that that's probably why I became an engineer at the beginning. And I start my career in networking during Cisco, joining Cisco when I was there at school. Then I go a little bit deeper. I start; you know, getting the CCNA certificate and go through CCNP all the way to CCIE. I remember that time, CCIE is a big deal. It's quite a challenge. I remember like spending all this weekend study. At that time, the CCIE labs, two-day lab, is very fascinating. Give you a very complex network problem, give you a whole bunch of equipment, you build it, hands out. I remember second day and during lunchtime, everybody had to leave the room, go to have lunch. Somebody would come in to mix out your network, the rest of that, and then it's troubleshooting. So, I thought that that's kind of like kind of related to fixed problems to solving problems. So, and a little bit not so intentional. And then later on, I get into security. It's also by accident, not my tough choice, but someone left the team, the team needs someone to manage the entire VPN portfolio, and that's how I how I get into security from VPN, and I get into the input security, and then from there, data security. But one thing I feel like I black, and I got into the CISO's office. So, it also by accident, not by choice. And I wasn't reported CPO, CPO left. I'm the only person working on the security residual. It's like, why don't you just go work for the CISO? And that actually put me a completely different perspective looking at security. Because I most of my experience is building product, you know, from the vendor side. But sitting in the CISO office, now I realize how much noise they got. And how to deploy the solution to it's very different than you know, from different sides. So that that is something that you know, not intentional, I'm so glad happened to me. So that's me and security. I think security is I think that that really gave me a different person perspective by looking at the security, right? So, I think that's it. So, like you know, maybe I never become the journalist, and but I just still have that journalist in me. I just change the subject, I investigate, and that's a brain. Oh, of course I have to talk about being a founder. That's also it's not intentional. So, I was looking for something I need in my previous job, I just can't find it. And it never really crossed my mind I will become a founder. At that moment, I don't know what got into me. I just feel like I can't I can't find it, but I really have a very clear idea what I want to do. And I really pretty much leave faith and think that, hey, maybe I should give it a try. If you ask me two years ago, three years ago, I probably won't believe you. If you talk to a lot of people in my career I met before, they probably would be surprised as well. But that's a life, man. A little bit of lack of a community and a little bit of what you want to do and where I am. That's my background story.

Jess Vachon: 08:57

That's awesome. And that's a lot. And I want to pick apart some of that. One of the observations I had when you were talking about going from you know being a journalist to cybersecurity is that there's a lot of things that that are very similar between being a journalist and being a cybersecurity professional. You know, you're trying to look at the facts in front of you, you're trying to see different vantage points and decide, you know, what is what I'm seeing true or not true? Am I seeing false positives? Am I seeing true positives? And that's a security instinct. You know, that's at some level it's threat modeling too, right? Because as a journalist, you want to know what you're putting out as truthful and accurate. So, you have to dig through all of that. And I think that's the same thing when you're trying to, you know, establish a security pro program or mitigate security threats. So, it's always interesting to me when I have guests on the show that say, I, you know, I don't really understand how it connected. It's from my perspective; I can hear when the guests are on how everything connects. I mean, you talked a little bit about your family, you talked a little bit about feeling like an outsider because of cultural and societal differences that you have, you know, I guess I would say amongst Americans, but and then I see that and I realize like all Americans are different, right? So, it's interesting. It's interesting. And then I think I hear a lot of the guests also say, I felt like I couldn't find my place, and then all of a sudden, you know, they're on the show and they're talking about the places they found. So, it's very, very interesting. You know, you talked a little bit about your childhood.

Jess Vachon: 10:46

I like to ask my guests, like, what influences did your parents have on you, on either who you are as a person today, or maybe those career choices? Because you know, we aren't just born and we become who we are. Our parents have a good amount of influence on us, and I'd like you to explore that a little bit if you can.

Jaz Lin: 11:08

Yeah, I think one influence is how different my mom and my dad, how different they are. And it I it's not a lot intentional. I feel like one thing about my family is that I think they always they always allow us to be different. Like I never give a crap like because I'm different thing, I've been punished. I've been punished is because I didn't do my job. Yeah, and my mind is very critical. I think maybe my mind and the family, we all have that kind of the outsider. Right? So, like I think that sometimes I always I always can find why this is the way. I always ask that question. And my mind's that way too. Maybe it's like I got from her, maybe because of the environment they give us. Like sometimes I have to remind myself, don't be too critical. Like ask questions, sometimes I annoy people. I have to read the room a little bit because I usually have a lot of questions. Not like around challenging you, I just have questions. Sometimes I have to read a room a little bit, ask me to die down, tone down a little bit. Like even with my thing, right? If you have 10 questions, just ask five. The other five, you ask separately. Like, I just that's just in me, sometimes like bothers my partner as well. Like, why do you have so many questions? Just like just let it go. So, I guess that is just like in our dining table, we always have dinner together. In the dining table is like everybody's like hands everywhere, try to try to talk about what's going on and always question each other. I guess that's like I guess that that's the main thing. And another thing is very important for my family is that you have to be responsible for yourself. I remember my first international trip. I already wore my butt off to save the money to pay for myself. And I'm so glad because I, you know, I mean, if I don't give something, I don't get upset, I get I don't get disappointed. I guess these will be the two things I would think that I remember I remember the most from my family, from my upbringing.

Jess Vachon: 13:21

That's awesome. So, you, you know, you grew up with a family, you moved out of the family. Tell us about that block when you're in, say, your early 20s or 30s? What was that like for you as you were saying, journalism, information technology? Talk to us a little bit about that discovery phase.

Jaz Lin: 13:39

Yeah, yeah. So, in my 20s and older to early 30s, I travel a lot. I backpack a lot. I'll find any opportunity to travel, and I write because I want to tell the story to my family. I start using block, writing blog. That telling a story is always in me.

Jaz Lin: 13:57

Like traveling, and I was always building some kind of movie, writing some kind of blog. Even now I just came back from my trip by building the whole like the movie one of shows with my friend. That's always in me. In my early 20s, I spent a lot of time. I find funny, in that part of my career, I've tried to find jobs so I can travel. So, I remember I was part of the marketing team with a lot of trade shows. And then then, like I guess I have to say that technology, I love talking about technology. I love to like to tell a story about technology, making the way that I practice on my mom. My mom has no technical background. I take that and try to explain to my mom. If she cannot understand, inside me, I feel like, oh, it's a big win for me. Even my mom can't understand. So, like, and also like traveling, also the telling story. So that's my early career. So, like I spend a lot of time finding a job I can travel. I really enjoy like, you know, talking about technology in different culture backgrounds. Like if you go to Korea, you know, you're dealing with different cultures. You go to South America, different cultures. That, I mean, that kind of experience, that memory is always stay with me. I just love it. Yeah. Do I answer your question? I feel like mumbling.

Jess Vachon: 15:18

Yeah, no, you did. Let's drill down on that a little bit though, because I think that's an important part, the traveling and experiencing different cultures. How does that inform how you think? And how, if at all, has it informed or guided your career?

Jaz Lin: 15:35

I didn't go a single thing. At that time, I'm working for very big companies, Cisco. Like in the in US and in Silicon Valley, we're always driving the technology. We are so comfortable to bring that, hey, this is the way it should be. But sometimes it's not that that's not how other people do business. They have different technical problems. So, we have to be created how to really introduce the technology that exists. We know how to solve certain problems; we can introduce it. You always want to be creative. How to use what you have to solve the local problem. That I love that. I love that part of my job. Yeah.

Jess Vachon: 16:20

Let's talk a little bit about intersectionality because you touched upon it earlier in the conversation. And it's important to me because part of the reason I have the podcast is so that I can bring diverse viewpoints and diverse people or people from diverse backgrounds onto the show. I've probably talked about it before. I am from the queer community myself, and I have my own opinions about what it's like to be queer in technology, but you've touched upon that a little bit. Why don't you tell us from your perspective? You know, have you seen a shift from say 20 years ago to now about how our community is accepted in technology and how we've come from maybe being out in the shadows to being leaders in the industry?

Jaz Lin: 17:09

Oh God, this is a tough question. So, I thought I see the difference, but things take a different turn. So, it's hard for me to answer questions because you asked me this question maybe five years ago, my answer is different. So, it's a lot of up and down. So, like that's why sometimes behind closed doors, when I spend time with myself, I always want to ask myself, what's my point of view? Do I still have that? Am I still the outsider within myself? Right. So unfortunately, I don't have a very good answer for that because I feel like a live background right now. And but to me, that's an environment, right? Me is still me, and I still feel comfortable when I'm behind closed doors with my myself. I think that's what I'm trying to focus on right now. Does that make sense?

Jess Vachon: 18:07

It does, and I think you answered the question beautifully. So, thank you for opening up about that. So, let's talk about you now, builder versus buyer, because you've got some things going on. Tell us about your business and tell us about your process or your pre your steps to becoming a founder and what you love about it and what's challenging about it.

Jaz Lin: 18:29

It's so challenging. I'm not gonna lie. So how I became the founder in my previous job, and that's I really appreciate to have that opportunity almost in the network security company for 911 writing space. It's very mission critical. So, there's an opportunity again involved with that, you know,

Jaz Lin: 18:49

AI, and I just realized that hey, I can't find a tool to really manage the AI agents, and I have to anyway have Amazon to do it, so have partner to do it. So, like I really, I see that coming, the huge wave coming. Now AI is so powerful, so optimistic, so much possibility. I it comes with cost. So, I quit I quit a cushion job, and it's not like I look, I see myself on the job. Now that I have a deep pocket, but I feel like I want to give it a try. Right? I know that success rates are low, but I gotta give it a try because I feel like there's something in me, I just cannot let go. And that just happened, I don't even remember now how that that's what happened. So, what I really want to talk about is this AI era, look, this is inevitable. Every 15 minutes in a conversation, somebody had to some somebody bring this up. What worries me the most is that we're so obsessed about speed. We're so obsessed about the speed, and everything has to be quick, and AI empower us to do something very, very fast. So, in my opinion, that speeds might not be the and security might not come hand in hand because if you want to do the thorough investigation, it takes time. Like if you look at today's startup world, like people just expect like a startup company or getting to the one million AR in one year, that's it happened before. But it's just it's crazy, it's insane. Right? So, and it's extremely difficult to get in touch with people right now because the app AI is so good to come on a pattern, process quite a large volume of data information and come on a patent and give you a tour, do think fast, and become very productive. But I don't want the productivity to be mainly predictable, you know what I'm saying? Like I still want me to be me. Like behind the door, I stand my own perspective, and not that your machine predicts me who I should be. But unfortunately, unfortunately, right now you open out the link and hey, I'm part of the well, and I use AI to beautify my writing as well, right? But a lot of content is produced by AI. I don't know how much this is thought to, right? And building our trust becomes so difficult. Everybody reaches a certain common conclusion, and everybody wants to fight for the speech, so obsessive about speed, including myself, right? Like sometimes speed becomes so counterproductive, especially in terms of security being relationship within trust. So, so the AI technology comes a lot of opportunity but comes with cost.

Jess Vachon: 21:53

Yeah. I share a lot of your perspective that you just brought forward, and it's interesting. I'm going to come back to journalism again, not in Intentionally, but because as you were talking, I was thinking about one of the things that I'm finding to be very sad is that a lot of the stuff I read constantly. I'm reading from the time I get up until the time I go to sleep. And I'm seeing so much of the content now where you can tell that it was written with Claude or it was written with Chat GPT because there's tells in the in the in the writing, there's common words that get used. And it's so disappointing to start reading an article that has some of your interest, and then all of a sudden realize, oh, this was just put together by AI. And to your point about how rapid the companies are trying to move, they're so focused on making the dollar using AI to justify using AI that they're just throwing the quality out the window. And to your point, we're starting to lose the humanity that goes along with it. And I've talked about this before. And I'm glad you brought it up as something for us to chat about. Here is that if we get so focused on AI and replacing everything with AI, what is there left for humans? Right? Because we're here for and we feel like we need to have a purpose. And if we hand all that purpose over, what's left? So, I'm glad that you touched upon that. But I want to drill down again on what you're doing, like how you know, in terms of security or even making the world a better place in terms of using AI, what are you bringing to the table? Why should people come to you and say, you know, so it's your product versus someone else's product?

Jaz Lin: 23:43

Yeah, yeah. So again, I want to touch, I'll answer that question, but I want to touch a little bit. I just such a mixed feeling. Like, like I feel guilty as well. Like I have to use AI to increase my productivity. Where's the line? Where's the line? That is a question I constantly have conversation with myself. Right. So now come back to what I'm building. So, I also want to ask you this question. Because the AI engineer right now is so focused on speed, I feel the gap between security team and engineering has become bigger. Because that the speed

Jaz Lin: 24:21

they try to build and the poverty the AI team, they got so much tool. I feel like there's a gap. So, I want to pause a little bit before I talk about line building. I want to especially like you are so deep in in security. Is that also why you see, or you see different?

Jess Vachon: 24:45

That's an interesting question you pose. So, in cybersecurity, I think we're using AI as a tool because we need to, because there's so much data we're trying to pull together. It's just it's not humanly possible to correlate all of it. So, it's a struggle because I work with my team and we talk about okay, how are we using this? We're using it, gather the data and have it give us some graphing or have it, you know, give us some suggestions of what it's seeing in the data, but you still have to be skilled enough in your work to know if that data is accurate and or not. And then you have to understand, okay, it's taken the data, it's manipulated it and given it to you in a way that you can understand it better. Now you have to make decisions based upon it. And how do you make those decisions? That's not something we can leave to AI. It can suggest things, but it doesn't live in the real world with working relationships like we have, which are built on trust and built on making sure that things are as accurate as possible. So, there's that element to it, and then in the broader IT organizations, I do see gapping. I see people who either are not learning AI skills, such as, and when I say skills, I mean just basic quality prompting, or they're thinking that they don't want to entertain it because it's going to replace them. And the and the sad thing is that you can cover your eyes and you can cover your mouth and you can cover your ears to AI, but it's coming anyways. And so, I think it behooves all of us in technology to find that balance in using it as a tool, not turning our whole careers and our whole lives over to it, to still think critically and interpret the information and still work person to person with what you're trying to do. I think kind of what you're getting at is do are we starting to see another technology gap like when personal computers first came out, right? Wealthier families had it, less wealthy families didn't, and there was a technology gap that happened. I think we're seeing something similar with AI, not necessarily based on dollars, but based on access and understanding around it.

Jaz Lin: 27:08

Yeah. Yeah. Yes. So, the problem that that I'm trying to I hope I can solve is that right now, like engineering are building so fast. I the gap that I'm seeing here is that they move so fast and they're so deep, but security team might not really have the visibility and the control to it. Like the AI what AI is a huge violator, and if first of all, what everybody means they all use AI, but they mean differently. So, let's start what we're building here is really like we're building visibility

Jaz Lin: 27:47

and policy platform for AI agents. So, security team will be able to have visibility and more control on what AI agent does. Talking to talking to builders, I understand that we are right now today we use a RAG, we'll provide instruction to make sure that hope that each agent will follow the instructions. But that probably is good for like I mean, just come out number, right? 50, 60 percent. I'm not sure how true it is. There will be, is any there will be certain things that in the grace areas is not acceptable. You have to be here to like these 10 questions. I just want to have a black and white runtime control. Like that, that I think that is still important. Right now, I feel like language banner becomes more and more powerful. A lot of time is still very arbitrary instruction base. So, and also can security really have a full visibility into it, have a full control. So that is the area like we are we are focused on right now.

Jess Vachon: 28:48

Yeah, and that's you're 100% correct on that, because I would, and I've said this before, security teams right now, we still can't see into the machine, so to speak. We can see activity around the outside of it, but we can't see if it's drifting, we can't see if it's breaking outside the parameters that are set for it. And we're having to work alongside or trust our partners in technology that they're you know kind of sitting in the saddle on the horse because we're running alongside the horse and we're not keeping up very well. So, it's yeah, it's a very important aspect. How is your tool addressing that? Like, how is it giving the visibility, but how is it also making sure that it knows that those parameters are being followed and not violated?

Jaz Lin: 29:34

Yeah, yeah. We're starting in the process of building it. I think the visibility is we'll try to gather all the app evidence to find. We're very laced to focus on, first of all, AI agents, when everybody's talking about AI agent, they might mean different things. Some people mean that the closeting in their laptop were really focused on the cloud-based AI, AI agents. So, I kind of feel like majority of SaaS company eventually become some sort of genetic company because a lot of SaaS-based tool flow can be automated by AI. So, these will be sitting somewhere in the cloud. So, and we want to make sure that hey, whoever using the third-party AI agent still can bring in their own their own policy. So, like the enterprise that they're so used to being your own key. How can we really make the job easier for the SaaS vendor? They can really have a customer-based policy, not just that everybody still shares generic policy. That's one thing. And then really that's one-use case. Another use case is that enterprises within enterprise, they start building the AI agent for the internal employees used to increase productivity. I think these are part, probably these are part, if I need a, you know, I probably want to see how many of them are there in the organization. So, and once we have the visibility and I probably want control, intelligence is hard, contest intelligence is hard. But first, can I see all the activity activities? Okay, you either 90%, 60% probably can be covered by how you build an agent, but it's gotta be something, it has to be black and white, that I can't have full control. So, these are the high-level two type of use cases that we're addressing right now.

Jess Vachon: 31:29

You said something and I that scared me, and I don't think you know that you said it or may not have heard yourself say it, but you mentioned individual customers bringing their individual policies into AI, and immediately the first thing I thought was if a vendor is trying to adapt their tool to individual customer needs and it's AI-based, what does that do to the model? What does that introduce to the mix that may throw these rules or controls that the vendor has built around the AI? What does it do to those? Right, and how because I I've been thinking about the model itself just you know, self-realizing for lack of a better term, or breaking out of the box. But I never thought about hey, you're having this human-to-human interaction, the vendor's trying to satisfy the customer. The customer says, hey, I have a policy that's out of the norm, this is what I want incorporated into how your AI agent works, and that's that change is made, but it's not tested. And I just thought now on a grander scale, we start to have 10 or 20 of our vendors in the environment who now have AI enhanced tooling, and they start to tailor the product to individual customers. What threat does that bring into the enterprise? And are we talking about it and are we thinking about it? And I don't even know that you know that you opened that door, but immediately when you said that piece, I made a connection there. What are your thoughts on that?

Jaz Lin: 33:08

Yeah, so I think we might make mean a little bit different, different, different things. So, like if I use in the third-party AI agents, but I don't want, I want, I still want my policy to be enforced. Not so much about the model itself. It's more like the AI agents, well, everybody, I one of the use cases when I'm working with one of my customers that they are sold into enterprise, but each of their enterprises has different neurons, right? So, how can you ensure, like, hey, certain things I want to match? The other customer might they might not care, but how can I really ensure, hey, I care about these four things? Can you make sure the same AI agent can enforce these four things? Because I know that you send AI agent have to serve 10 different customers, but this is the four thing I care. Can you please make sure these four things is black and white? Not like, hey, you know, yeah, that's what I mean.

Jess Vachon: 34:11

Yeah, yeah, and I and I see what I see what you're saying, but I looked at it from to your earlier comment. I looked at it from a completely different perspective, right? I looked at it from okay; I'm more of a on the purchasing side where you're talking about providing that safety and security for someone who's bringing these tools into their environment, which is I think that's the builder versus the buyer mindset or difference. And yeah, it's very interesting. So, it sounds like you're really focused on that assurance piece for people that are bringing these tools in to either enhance their existing product or their operations. Is that correct? Am I dancing close to the closer to the floor?

Jaz Lin: 34:56

Yeah, yes, yeah, it's too dependent use case.

Jess Vachon: 35:02

So, we're talking about the here and now. Where do you see things, or where do you see your product, or where do you see security for AI in three years or five years?

Jaz Lin: 35:16

Wow. Be honest with me, I don't know. The market is so noisy right now, and things are changing so fast. And I think right now a lot of still in the air. A lot of agentic application right there right now right now is still good enough, good enough. So, it's

Jaz Lin: 35:39

good enough, good enough that I don't, I'm not sure. Right. So, I think two years a long time. I we all lost out, we all lost out crystal ball. But I do think that I think that policy should not live in the application. It's it has to be separate. I still in that you know school of thought, I feel like I would prefer to decouple the policy from the application, put a policy in the cloud infrastructure or some kind of infrastructure. I want to be able to separate that. And I think, and also like you have to be flexible enough to resilience based on all this crazy thing can come out, all the good things can come out. Language model might have another lethal moment, right? So, how can you adopt the all the best plug and plane, make sure you're flexible enough, and the and the application is application, you still need to have your cloud infrastructure to be able to address all the different things. So, I still prefer that to decouple the policy fund application, putting the infrastructure.

Jess Vachon: 36:48

Yeah. Nice. All right. Well, I think we've drilled down as far as we can get on AI right now. And I think everyone shares your view. Like to say, to ask the question, and I kind of asked it tongue in cheek. What do you see things three to five years? We don't even know where we see AI in six months. And that's, I think if we're being truthful, we're all just kind of holding on, trying to hopefully see that this is going to end up someplace rather than continue to be something that we none of us feel like we can wrap our arms around completely at the moment. So, your founder, that's stressful. That's being always on a lot. What do you do to relax and recharge and come back fresh?

Jaz Lin: 37:32

Yeah, so I think I really like running. Running and cycling. I think I like running the most. And it's easy, you bring just your running shoes. You can do anytime anywhere if the weather permits. And also running is also the way for me to get to know different cities. So, every time you go to travel to different cities, if I can, I

Jaz Lin: 37:58

always like to go running in the neighborhood to see, you know, I don't know. I just get to know the city better when I when I run. Try to keep up, try to keep up right now. You know, I try to remind myself, hey, I could get to keep on my running schedule.

Jess Vachon: 38:16

Yeah, nice. I I'm a runner as well, so I certainly can understand that. And I agree with you. I love you know when I'm in a different city to go out and run, because you get again, we're talking perspectives so much. You see the city in a different way. And I feel like you become a small part of a city or wherever you're visiting when you're out running, because you're it's very intimate, you know, it's foot on pavement and then being deep in your thoughts. So, thank you for sharing that. So, you know, we're coming to the end here. Where would people find you? And I mentioned earlier that you're still writing. Where can they find your writing? Because you do a lot of writing. I've seen it.

Jaz Lin: 38:58

Yeah, thank you. Yeah, so I think Linkin's probably easier, easier ways to find me. I also had a median. You can look for me, Jaz Lin at Linkin, and also have my Medium account, medium/JazLin. So, and you can find us while we're building as Skyrelis, which is www.sky r-e-l-i-s.com.

Jess Vachon: 39:26

Awesome. All right, just before we close, I need to say a word about our new partner for the podcast, someone who helps make this show possible, Onyxia Cyber, the operational cyber resilience platform transforming security data and asset intelligence into action. If you go to onyxia.io and you mention this podcast when you talk to them, you will get the first month's service free. So please reach out to them. They're a generous sponsor of the podcast, and they allow me to bring great guests like Jaz Lin on the program. So, check them out. That's the episode. Thank you, Jaz Lin, for the stories and for reminding us that the angle you're standing at changes what you can see. And to everyone listening, bring your whole self to work. It's the part that they can't automate. Until next time, stay vigilant. Bye, everyone. Bye.