Inside Artemis' "AI vs AI" war | Shachar Hirshberg & Dan Shiebler (Co-founders, Artemis)

Josh: Shahar Dan, welcome to the show and congrats on the launch. just to make sure everyone has enough context, could you start by explaining what your company Artemis does?

Shachar: Sure. first, thank you so much for having us for, for all the support throughout the way. Uh, we really appreciate that. Um, so Artis helps companies detect and stop attacks in their environment across all their stack, their cloud identity, network information, everything that could try to target their organization.

Artis detects the bad things and helps companies stop them before they can impact their organization.

Josh: So you've just come out of stealth, you've announced your seed and your series A funding along with your first several customers. Can you give us a sense of where the company is right now, team size customers, what the product looks like in production?

Shachar: first, because we grow so fast. Everything changes every week. So. in the past, so we started seven months ago and we are now about 30 people in the team, um, growing about one to two people, every week and scaling across engineering, product research, and go to market.

the product that we built was built in a nine 80 way from the get go, and I'll let that data, the technology of it, but just from the product perspective that allows customers to get a dramatically better value and ease of use compared with traditional legacy SI and legacy detection products. And what we see in practice is that. The users of the product spend about three to four hours every day, in the product because they're just able to get the outcomes they want using Artemis and most security product. It's kinda like a box that sits there and they don't really know, what is happening. so we are really happy about the engagement and obviously work with the customers to make it even more delighting.

Dan: Generally what we see is that, that we've been able to move very quickly because we've built the company in a way where we can develop everything entirely with AI native tooling, we've developed our code base and developed our internal processes so that everything is entirely leaned into what AI systems are capable of today and where we think they're going.

Continue moving in the, so there is no, traditional, Manual work required to be able to bring concept into prototype and bring prototype into the hands of customers and iterate very quickly. We're able to iterate on customer feedback extremely quickly, and also iterate from the things that we see in the data into improvements in the product itself in a way that's both secure and very fast.

Josh: Shahar, you got to know. Customer and the product very well from your time at AWS and Palo Alto, in what ways is that an advantage, and in what ways, if any, is it a disadvantage when it comes to starting a brand new company in the space?

Shachar: Yeah, that's a great question. So I've been in this space for, in technology and cybersecurity for the past 15 years, but in the broader security operations where we operate for the past decade, I kind of seen all the, I'd say two, two last three iterations of this market. And this has been particularly valuable because.

I know the customer very deeply. I worked with them on multiple products to help them achieve their goals in this area. And that builds a lot of trust and also a lot of intuition on what customers need and what they need and don't even know they need yet. But they will discover that in a few months.

And that's allows us to be on top of things in terms of the product roadmap, which as Dan mentioned, we just deliver on very quickly. I think the flip side is that you are seeing, because you've seen different products in the market, you know how things are. So you have to continue to push yourself to reimagine what is possible today when you rebuild everything in a native way.

And this is why we've been really pushing hard on ourselves, but also everyone on the team to adopt every AI native technology. Use it. All the time in every possible ways, and then bring the learnings back into the product. And really every paradigm that we think that customers need, we go and look at it from the lens of, okay, because we're starting from scratch with AI being so powerful today, how we can reimagine and make it a hundred times better compared to what you used to be.

Dan: Security operations teams are dealing with a lot of the same problems they've been dealing with for a long time. A lot of things have gotten harder and more difficult as the landscape has shifted.

Josh: Like what type of things?

Dan: both attackers have much better tooling today and can move much faster. So it raises the bar on what kinds of response are needed, and software has become much more complicated.

And that also has raised the bar on how difficult it is to secure states. the tooling landscape has changed and what people are able to do today. What's possible to build today is changing, but a lot of the underlying core problems are either the same or the extrapolations of the same problems that have been there for a long time.

Josh: So Dan, you. Were an ML engineer at Twitter before you went to Abnormal. Are there things you learned at Abnormal that you're applying here?

Dan: Yeah, absolutely. The, a large part of the goal at Abnormal is building a product that is able to understand behavior and then utilize that understanding in order to detect cyber attacks with extremely high precision. This is a very core element, both in abnormals products and in many other, cybersecurity products that are really leaning into this kind of behavioral based and AI native architecture for being able to provide value to customers.

what we're building is in somewhat of a different way, and a little bit more of, geared towards a slightly different suite of problems, but leaning into the same kinds of core underlying technology and the same kinds of, overall objectives of building something that's able to utilize an understanding of a customer's environments in order to be able to detect cyber attacks in a way that's resilience, to changes in attacker behavior, and requires as little manual work by the customer as possible in order to get that defense and get that protection.

Josh: So let's back up a little bit to when you were getting started Shahar. You knew you wanted to start a company, and you spent a lot of time sorting through ideas before you were ready to leave AWS. what did that period look like for you? How did you, approach the idea exploration phase? Are there questions you asked or things that helped you either rule out opportunities or lead you in any direction?

Shachar: One of the things that I focus on and still today is working backwards off customer problems. So when I looked at the ideation process and a large part of it was just Dan and I spending a ton of time together thinking about the future, having fun building and chip. But, a lot of it was thinking through what problems we know, talking with hundreds of security professionals to understand how what they feel today might be amplified or reduced as the word changes with the introduction of our technology and a lot of the ways that we operate is okay, how can things change if AI will be 10 times better next year and hundred times better?

we spent a lot of time working for ideation and answering such questions. I also participated in first rounds PMF method program, which I highly recommend and thank you for making this reality. do you want, by the way, to share a bit about what the program

Josh: Yeah. Why don't you share a little bit of, from your perspective about what the program was?

Shachar: and I think I even posted about it on LinkedIn, is like the four most valuable days that any founder can get.

Josh: magic to my ears and definitely my partner. Todd's ears.

Shachar: And it's, like equity free, everything. Free, actually everything free. This, that was worth PET was great as well, but, program

Josh: economics, it's just, it's just programming.

Shachar: exactly. So just thinking through company building, thinking through how to ideate and what are the first steps you need to go through in order to get from zero to one.

So that program was, incredibly valuable as well. And then. As we, let's say, converge in our thinking. We honestly came back again and again to this market of security operations because both, I have been in this market for a long time and Dan is work on detection problems for the past really 10 years.

And you wanna add on this?

Dan: I, I mean, I think ultimately at the end of the day, you need founder market fit. If you're going to start a company, you need to be working in a space where you really feel that you have a differentiated understanding of the problem. I, I think that there's, there, there's a lot of very large problems that, uh, people have, that companies have.

There's things that you see growing in the future and we saw a very, very large problem in security operations. We saw a problem that was growing, growing, and we saw an area that we had a really, really deep understanding of how this problem is shaped and what the different kinds, what sorts of solutions work, and what kinds of routes are dead ends, and had confidence that by leaning in and if we are able to get the right, people to join us and able to set the right, uh, core initial standards, we could have a really good shot at tackling this problem.

Josh: So you're talking a lot about the we. How did the two of you decide to become a, we? What was the co-founder dating process like?

Shachar: Yeah, so Dan and I, met a few years ago in New York. we got introduced via a mutual acquaintance and, it, it was actually a pretty funny story because, we basically scheduled a coffee. We're like, okay, let's, meet for 30 minutes, get to know each other. And the whole context we got was both of you guys are in security.

New York, just have fun. And then I think the first time we met, we spent like probably four and a half hours just walking around in the park talking about the future of security, talking about how things will change, what problem people can solve and all of these things. And, it just continued from there.

We spent a ton of time together ideating through problems and projects and building together and some side projects. And, it's been an amazing journey so far.

Dan: I, I really feel like one of the, one of the things that stood out to me is,the depth of Shara's commitment. and in terms of really wanting to build something great, wanting to see the path to the future and choose the right path. and I think that's, that really is something that,was exciting to me because I feel it's really crucial that, two people are on a journey like this together, have this sort of shared commitment.

I think that builds on top of our shared chemistry and the way that we compliment each other. but I think that kind of, mutual goal and mutual objective and overall drive that we, we see in each other and compliment each other with really is what was very clear from the very beginning of the partnership and was strengthened over time as we explored lots of different angles and discussed lots of different ways.

Josh: You were both at working elsewhere. When you decided to take the leap, did both of you think it was the right time or did one of you jump before the other? How did you know when it was the right time to say? Okay, we're done talking. Now we wanna start building.

Dan: We were both pretty aligned. The, there had some slight differences in terms of exactly which month, made sense, but we were, basically we picked, okay, this is like the right time period and this is the right opportunity. And, I think things ended up moving a little bit faster than we expected.

And then we ended up being very happy about that because of the, opportunities that were opening up in the market. I think that in, in retrospect, if, if we'd waited an extra six months or, or so as we were, were, originally discussing, then probably we, we may have ended up being too late, uh, given the, given the shape of the market, given the speed in which things are are developing.

we wanted to make sure that we had the confidence. I'm also glad we didn't start the company a year earlier because there's so much different things that's possible in tech today. We were able to build in such an AI native way because of the moment that we did start.

And so I think it was a combination of us being strategic and us being pretty aligned and really having the commitment, there's the right choice to do this with each other.

Shachar: And I'll add, there is never a good time. So it's always, always have like more promotions, more money, more stuff, more things can happen. but we just saw the pool from the market, just how customers really excited about the problem we're solving. We saw how it's changing and we realized that we have to go right now in order to fully capture this huge opportunity.

Josh: Let's talk a little bit about hiring in the AI era. are there any startup norms of hiring that you're rethinking

Shachar: Yeah, so first people is the most important thing in the company. Without people, we don't, we won't have a successful company. So we put, I would say over 60% of Dan's time in my time on hiring and developing the best talent in the world. And we try to make the hiring process extremely easy and fast for candidates.

Josh: what does the interview process or hiring process look like?

Dan: first. One thing I'll say is that we weight very heavily on references. we put a lot of emphasis on speaking with people who've worked with somebody in the past, and we try to, front load that in the process and move that through very quickly.

Josh: And you found that to be a good predictor?

Dan: we have, we found that to be extremely indicative that really we've, everybody, I've thinking we can plot out on an al almost perfect correlation between the strength of references and the strength of which, somebody actually delivers.

And I, I think also we don't aim to do. it was super lengthy interviews. We usually have about two hours of in-person meetings with the team in addition to meetings with me. And, Shahar, we try to do as much as possible in person, as much as possible, as, as fast as possible.

Josh: Do you give a love any work projects?

Dan: we've experimented with it, especially when, we, we did one experiment where we hired, winter interns, which ended up being a very successful experiments.

And we did utilize work projects primarily because of the sheer number of applications that we got and the, there's always a lot more. Risk that you take, bringing on somebody who doesn't have substantial work experience bringing on a student. So we wanted to, to de-risk that. And what we did with work projects, in that case is we didn't focus on, asking somebody to complete some coding task and then review the code.

We asked somebody to send us a link to a particular part of the, a website that they had built that accomplished a particular test, never looked at the code, didn't look at the implementation, didn't care if they built it with a no code tool or coding or cloud code entirely. It's just you can, you, can you build something that was really what we were trying to assess, in, in that work project.

And I think that's in general, I think that the, that one of the big questions we want to ask when we're hiring somebody is, are you capable of building something from scratch? Basically anybody we hire wanna have that capacity.

Josh: How do you assess for AI fluency, in your interviewing process, just given the paramount importance that you're.

Dan: This has changed over the course of time that we've been running this company because this has been one of the largest changes. When we, when we began this company, there were, there were some people who were extremely in, but it was a much smaller percentage of the market than today of this has been, of course, a absolutely crazy period of seven months in terms of a AI adoption of generally somebody who is a true builder and who is open-minded and who is willing and excited about utilizing the best tools and up-leveling their capabilities, whether or not they've had the opportunity to.

Really lean into AI native work. Some people work in companies that's forbid a AI utilization in their work. And if they have experience with ai, it's only in in small side projects. Other people work in companies that are extremely AI native. I, so I think that the, where somebody stands in terms of their prior experience at this point right now, given how long of a time that these tools have been around is, is not necessarily a great predictor of how much they'll lean into it and really adopt it.

We do a very good job of, inculcating people in the very, in the first few

Josh: So are there specific ways you've set up the engineering team or the code base to take advantage of this?

Dan: Yes, this is, I would say one of the largest. Priorities we have in the technical architecture that we have and also in, in the team. so the two things that I would say are, one is that AI coding tools enable engineers to be able to reach over the line and understand code that's outside of the range of things that they would normally be able to understand normally be able to work with in a really unparalleled way.

That really changes the physics of software engineering. It becomes so much easier for people who have never had experience with writing terraform code, writing frontend code, be able to make changes, see things debug code, debugging, I would say is one of the biggest areas we see the most, Opportunity where it's possible for agents to be able to do really effective understanding of where are their issues, where are their problems. And so giving every person in the team the capability to utilize these tools and have that be the first place that they go, then puts everybody in the mindset of, if I wanna solve this problem, I don't try to solve the problem manually.

I try to figure out how to change my approach to setting up these tools, to utilizing these tools to make it so that the tools have the right answer. And that then goes into how we actually set up our technical architecture. When we make architectural decisions, the first question that we ask is this an architectural decision that will increase or decrease the capability of AI tools to have the, to make the right answers?

Anybody who's worked with AI tools knows that they often are wrong,that is one of the reasons why they're difficult in some capacities to use, especially in large and complex code bases. And there's certain dis connectable decisions that you can make that make the tools more or less likely to be correct.

And we've structured our systems to make that. So every time we make any decision, we're trying to make them more likely to be correct.

Shachar: People side, we were. Work to give people the best experience. So this includes typically wrapping up the full end-to-end interview process in one day or two days max. So people come, they have a chat with me or Dan, and then they meet the other engineers for a couple of interviews. They typically stay for lunch with the team.

'cause we want people to get to know the people that they work with and see, it's are, is it the group of people that they'll have fun working with and build with? And then, as Stan mentioned, during the references early, so we can just wrap up things. And we typically give the offer between like two days from the moment we started talking with you or not. and then on the last bit on the ai, I think we, at this point in time, every person we interview, we ask them, how do you work with ai? Everyone is like, oh, I work with AI all the time. But then we go deeper. To this day, we still haven't found someone who, before joining Artemis, has worked with AI as intensely and thoughtfully, like we work in Artes.

So today everyone is basically having like, let's say four to eight cloud code instances running in parallel, building the system, working on shipping for to buy features simultaneously. And there is a lot of art and science to it. And we teach it when people join the company and we look for the people who wanna learn and be at the cutting edge of these capabilities.

Josh: What about outside of engineering? Are there workflows that you've found personally to be the most impactful?

Shachar: We try to automate everything we had as a additive company. I think a lot of the processes are helpful because, or automating these processes are helpful because we're now 30 people. We were 20 people about a month a bit ago, and we were basically six people in September. So with these rates of growth, the company changes dramatically, and we use AI to make sure everything is still streamlined and working.

Then keep track on all the business metrics and the operational metrics without having five people manually tracking it and instead just using AI to automate such things.

Dan: I would generally say monitoring is the area where I've seen the most value outside of ai. The ability to be outside of value from ai, outside of just engineering directly. really the ability to see all the different things that are going on, within the company. whether this is things like chats, tickets.

Actions being taken in different platforms. And then certainly the behavior of the code base itself, behavior of how customers are utilizing the products. All of these different areas are things where traditionally, if you try to do data analytics over the different systems within your company, it is difficult to be able to know what are the right questions to ask, what is the right way

to

Josh: do you have dashboards that, like, are there metrics that you're looking at now that you never would've looked at before,

Dan: Some are metrics and numbers, some are insights and, messaging

Josh: such as.

Dan: things like, there's a, this is the tenor of this conversation. This is the, kinds of activities that are being performed right now. There's, questions you can ask about what are the types of activities that people are taking, what are the kinds of, Work that people are doing and how people are spending their time, that there's, we're experimenting and exploring all of the different things that, to some extent will probably be incorporated into dashboards and viewability. But it, I think this is gonna be a really strong backbone of how we'll be able to scale.

Shachar: I'll add one piece there. For example, in the product analytics, we look at how customers use the product and you could do it manually and I can watch recordings and have a data analyst going through the metrics. Uh, we use agents for that, and that allows us to understand where customers struggle, where there is friction, and automatically think through how we remove that friction from the product or see how we. for example, if they have a flow and it takes them two minutes, how we can add a new feature that reduces this to three seconds and all these things, you could do it before, but it was a lot of manual effort to do it at scale. And now it's largely the insights part to Dan's point is largely automated.

Josh: Let's talk a little bit about customers. How did you decide, you had a lot of customer disc discovery calls. How did you decide what customers you wanted to go after, what customers you were deliberately going to ignore? like, let's just talk about how you landed on the ICP that you're, executing against.

Shachar: The core problem for us is to, or the core, tenet for us was working with people that have a problem because the ICP we ended up planning on is largely upper market and enterprises that have used or are using a traditional sim, which is, basically security a. We got a lot of interest from mid-market downmarket companies because as a part of the exploration process, we wanted to make sure we were hitting all the different ICPs, but it was more so something they wanted to try because it sounds really attractive, but I didn't have a problem. We didn't have a pain when you speaking with them.

It's not like when we were speaking with the qcp, which tell us, yeah, my team is chasing 10,000 alerts every day. They don't really know if we have detection coverage or not. All these things where you can really feel the customer being in pain. So the prioritization for us was targeting the segment that is hurting the most, delighting them, and then expanding to the broader, part of the market.

Dan: There's a certain point at which a, a company's scale and complexity reaches the state when the need to be able to monitor all of this data and the difficulty in doing it in a way that has low noise and really gives you the, the insights that you need in order to be protected, becomes extremely challenging.

There's, it's a, it's really an inflection point in terms of the complexity of the problem, and we, we found that our ICP really sits on, on the side of the companies that have, are past that inflection point in terms of complexity.

Josh: How tech forward are the customers that have that pain point? Are there commonalities where you can generally detect the prospects who are gonna be experiencing the most pain right now? Or is it just a lot of conversations to assess pain threshold?

Shachar: I would say the pain exists in that segment that is covered typically over like 2000 employees or 1500 employees. And then it depends on the industry and the capabilities of the security teams in order to understand how much pain they feel. And they want to, let's say, outsource by buying a solution, versus building on themselves.

And we tend to see about 50% of our customer base is highly regulated industries such as financial services and financial institutions, where they have the need, the impact can be very high, but they don't necessarily have the, let's say. Desire to build in-house. Whereas we work with technology companies and we definitely have a sizable amount of my, of our customers as technology companies, but some of them prefer to build it themselves because they have engineers that work as security people, but they're in practice software engineers.

and for this segment we do serve them, but mainly because they come to us to use our MCP server to still build themself, but use Artemis as an additional source that help them to get dramatically better results.

Josh: Every founder remembers the first customer and what it took to get there. What's the story of closing your first customer?

Shachar: we closed the first few together simultaneously, so it was about the same time. But I think the key thing for us was reaching the level of trust and reaching the level of satisfaction where they actually came to us, the first three customers and told us that they want to buy before we ask them to buy the product.

So it's,

Josh: piloting.

Shachar: yeah, they were design partners, they helped us build the product and then they said it became such a core element in my security operations and capabilities that I want to make sure I have the enterprise SLAs and that I have the reliability of a product that I bought, which is obviously, no, you can't wish for anything better than that when the customers come to you and ask to buy.

Um, but, uh, the, the road there was, uh, long and, um, Very meaningful and we really value them. They've been true partners and amazing, thought leaders in building the right products that now many more customers enjoy.

Josh: Any non-obvious or surprising product market fit early signals that, sort of let give you the confidence to let you know you're heading in the right direction.

Dan: We've seen a real inflection in utilization there. there's one pattern that we see very commonly with the customers that work so far, which is they'll connect a couple of relatively easy to connect data sources into the platform. We will deliver an initial set of insights and they will add a bunch more people from their team and add a bunch more data sources immediately.

And that's, and then

Josh: So it's really the time from that first quick connection to that second where they go to a deeper level of connectivity. And that's the unlock that says you found something for them.

Dan: Yes. it, you have to earn that trust. There's generally, when you're working with secure teams, especially at a large enterprise, they have the permission and capability to connect you to some of the data sources that are important for them, but not necessarily all of them. They may need to get internal buy-in, or they may need to be able to get permissions that someone else has or they might not necessarily have in order to access certain very important data sources.

it depends on the organization themself, but in order for them to have the confidence to spend that political capital, you need to earn that as a company that's gonna be working with them. You need to show them that this is actually something that's really valuable. It's worth it for you to go and put these, spend that political capital, get the, these connections and put everything in because we're going to provide you so much value when you do that.

And we, we earned that by. Making things as easy as possible and showing as much as we can.

Shachar: It around. So we started building in late September. Around late January was when we, or maybe early February, was when the product reached a maturity of a platform, and that's when we saw things like fully clicking and we just saw the usage metrics, like how much time people spent in the product, what are the activities they taking with it, how many models of the product they're using, just so that like skyrocketing.

So the, the real, graph of has gone, basically parabolic. They were like, okay, this is working for customers.

Josh: So you started off with founder-led sales. what did you learn about yourself and about that process? And then how do you see yourself, growing your go-to-market efforts?

Dan: one thing that I've learned is. And, it's a lesson that I suppose I'm still learning every day, is that when you want something in sales is about asking for things, that's you need to ask for things. You need to ask to book the next meeting. You need to ask to get the introduction to the team.

You need to ask to, get the initial connection. you need to ask, and you need to give people value and give people trust and learn and listen. But ultimately, you need to make the ask. And that's, that's something that I think I understood to some extent, but didn't really fully appreciate until going through this process of founder-led sales.

Shachar: I think for me, billion Dan's point was a lot about trust. when I was in aws, you the default, right? So people will buys first and if they don't like it, they'll go but. With Artemis, it's about getting to a point where you're really more of a trusted advisor to the CISO or director of security Operations or even the security analyst that you work with.

And in this point, I think I'm on texting basis with every single customer that we have. And there is also a direct correlation, like, okay, if I'm not on taxi basis with that person, with that customer, probably a deal is not gonna happen because we haven't earned the trust yet to get to the point where they put Artis in the core of their security operations.

Josh: Now, does that scale, are you, do you expect that a year from now, two years from now, five years from now, you're texting every customer? Or how do you expect to grow to your go to market efforts?

Shachar: Yeah, I think I told them half of my day now is just texting like customers or internally. but uh, we just hired our first, um, um, initial go-to market hires, both, uh, account executives, sales engineer, and go-to market engineer. And we're now building the repeatable go-to market motion. And a lot of that is distilling the intuition that we have from the hundreds of customers calls we've done so far into a sales script and let's say clear en enumeration of benefits that customers get when they work with Artes.

And we'll see, it's the start of the experiment. I'm sure we will learn a lot over the next few months on what works and what doesn't work. And I'm sure we'll see the initial conversion rates go down dramatically because people don't know, or people that just joined the company dunno, Artemis as well as we do obviously just yet.

but, we hired amazing people and have a strong conviction that we'll be able to work with them and with the customers to ensure we're able to communicate the value in a repeatable way.

Dan: just to answer your question earlier, I hope that five years from now, our customers feel comfortable texting us directly. If they have any issues, they have any problems, that they have easy access to our phone number and text us day and nights. I think that is incredibly important to have that level of the ownership over every single customer that works with Artemis.

Shachar: One of the conversations we had with one of our first customers, he, he was basically saying that, he was like, I know now you give us the personal attention because it's your founder led, but as you scale, it'll be some customer success person that handles my request. And we told him like, like, like, no, you should, if anything is wrong, you should tell us.

Like, we truly wanna know. Like, it's not a you, us, it's not a distraction for us. It's not the news for us like customers and your happiness is the single most important thing that we are focusing on. Because if that works, everything else will follow.

Josh: So you, you've been pretty heads down, focused below the radar. Are focusing on building and just now as you announce your financing, your first customers,you're beginning to pop up. What advice would you have to, for other founders, especially AI centric companies, to figure out when to launch and when to open up their story?

Shachar: for benchmark of cybersecurity companies who are actually going outta stealth very quickly. Typically, companies stay about a year to two years in stealth. we believe that what we have built is differentiated enough such that we. Want to put it out there. We want to educate the market. We want to help customers secure their environment.

SAI is allowing adversaries to also accelerate their attacks. So us it was putting ourself out there so we can defend customers, and we just saw the traction from the market. We saw that customers started reaching out to us while we're still in stealth, which is extremely rare for cybersecurity. and that is because of the good word that our first customers have been putting out there and truly appreciate them.

So for me, the, the advice I will have for other founders thinking about when to go out there would be on the readiness of the product, because you don't wanna be out there and disappoint people. You should be a bit at least uncomfortable about the product, but not like, this is just not to work like it should work and deliver on the promise you give customers, even if it's a small part of it.

and second, ideally have some customers that willing to put their logos on your website or their quotes on your website to show the public support and share the value that they got from the product. Because it's ultimately all about trust. And especially now when building software is easier than ever.

Everything is just so noisy. So you need to stand out somehow.

Dan: I think that ultimately you don't need to go out stealth to delight your first customers. You need to stealth to acquire customers and. Be able to improve your ability to hire engineers, and

Josh: hired 30 plus folks while in stealth,

Dan: yes, we have.

Shachar: But most small companies are not able to do that.

Josh: so what makes you so good at it?

Shachar: okay, so one thing is our interview process, which a lot of the top talents, they always have options, right? And we try to make their choice very easy because we try to give them the best experience of what it's like to work in Artemis and Woo very fast. So while other companies, their process will take three weeks, they already have an offer from us after a day and a half.

And then we just might make that decision as easy as possible. And then secondly, we have been able to communicate to the prospects people. Chatting with how fast the company has been growing, both in terms of commercial progress and product progress. And this is something that is very exciting for people.

And lastly, I think people connect with our mission, which is helping companies defend themselves in the era of ai and they want to do something that is truly meaningful to others. and that has been helping a lot.

Dan: We make it a real priority. We recognize the incredible importance of having a team that is extremely top-notch and able to be in place and support our customers, and we really lean into that. We're very intentional and thoughtful about the process that we put into place and the experience that we give to candidates.

Josh: Intentional about the culture. If I'm, I believe you're building an in person culture here in New

York, has that helped hurt in.

Shachar: I think it helped. it's a self-selecting process, but we actually have a few people that were remote prior and they're missing human interaction. They're like, I want to be in the office. I want to be with others and brainstorm and whiteboard and just build together, which is very fun in early stage when you, it's all just greenfield.

so I think that was net positive.

Dan: I really agree. I worked remote essentially from. The last five years, before, before co-founding Artemis, and I saw how it could be something that makes it more difficult to really form strong bonds with your coworkers. And I think there's a lot of other folks who feel the same way, especially in New York where it's not the same as in, in an area like the Bay, where I think a lot of people don't necessarily live within the city, and it's a little bit more difficult to commute in.

A lot of people in New York are, there's a very good commuting network, so it's a maybe be a little bit less of an ask, for people to come into the office. And a lot of people crave that interaction and want to have that option and have the ability to form those bonds to their coworkers.

Josh: Besides for just being in office, it seems like you've also been very deliberate in the type of culture you wanna build. What was most important to you as you really, as the two of you said, spent time trying to think about the culture that you wanted Artemis to have.

Dan: We had a, I, I remember one of our, one of the meetings that we'd had where we originally wrote out our first set of company values, uh, which, which we iterated on a, a couple of times. And, and really I, I think have, you know, landed on something that we, we really feel strongly about. And at the top of the list is customer obsession, which I think is, is a core part of our, ethos as a company. but very, also very important is ownership. I think that it's crucial to have each person really exude ownership. That, and that breaks down in a couple of different ways. But a core framework to, to think about this when thinking about a candidate and thinking about how they fit into the company as a whole is autonomy as an element of ownership, which I alluded to earlier, building things end to end of really having the desire to own things and not throw things over the wall or say, this is not my job in, in the process of building something.

And another one of our core values we really select for strongly in the interview process is intentionality. we are very intentional about the decisions that we make and how we, we structure the company and the, we expect that people who join are intentional about the decisions they make and how they do their jobs.

So that, and how, people make decisions around the trade-offs that they make, the decisions and things that they build. And I think that people appreciate, that, appreciate working in a culture that, that prioritizes those things.

Shachar: We were also thinking a lot about what will make a company successful today and what will still allow a company to remain successful five years from now. And for example, two of our values, additional values are velocity and keeping the high standards, which oftentimes contradict. But in the era of ai, when you combine it with customer obsession and ownership, it's actually possible because the time to build went down dramatically.

And we look for these people who exhibit these qualities in order to deliver the best outcomes for our customers. And. We started with, uh, five company values. We're trying to keep it, uh, up to five because we, uh, want to make sure people actually know what the company values are. And one of the things that we do to reinforce it is that every weeks in the whole Hands meeting that we have, we give shouts to people that exhibited the company values, whether it's customer ion or Velocity or any of the other ones.

And that helps to really reinforce that these are the things that we care about and we as a company prioritize and value.

Josh: How often do the two of you disagree, and how do you work through that?

Shachar: I honestly don't disagree a lot. Oftentimes I'm like, okay, should we argue more? Like, should we argue at all? I think Dan and I share very. A common set of values and way of thinking from first principles, and that really helps us to just analyze the situation, get to the right conclusion, and think through what makes most sense.

And then we find the company, each person has their own responsibilities and like final say on areas of focus. And that also helps to just like streamline because we don't

Josh: Yeah, that's what enables the velocity goal.

Dan: I, I think generally. We have a very strong shared trust. And so when there are certainly situations when we'll disagree and discuss and come to a decision. but a lot of the times when we disagree on something, it is because one of us has thought deeply about something.

The other one hasn't necessarily thought deeply about it. And then whoever has thought more deeply about it, the person who has thought less deeply about it, just refer to them in most of the case, say, okay, you know what? You're probably right. You thought more deeply about this to me. And so that, that ends up leading to, a relatively, low amount of disagreement because, and that's directly derivative of trust.

Shachar: And most decisions are really two-way doors, meaning like we, and that goes for the company. We, with the other people, we ask people, okay, the decision we're making now, is it a one-way door big architectural choice that we will really need to invest a lot of time to change? Or is it something that we build sheep measure and revert or judge course?

And that allows us, again, to do both the velocity and intentionality. Like we definitely want people to think through trade-offs when they're making a decision, but they should make a decision quickly and reflect on it Also, after, in order to make sure we made the right one. This is also how we think about many of these things.

Like most things, it's like doesn't, they're very local. They're not a global decision.

Josh: So overall, the positioning you've landed on for Artemis is AI battling ai. Where do you think we are now in terms of how AI is changing the market and risk landscape?

Shachar: So part of the reason we started the company at the time that we started was because we have seen the freight landscape change from our purview at AWS and abnormal security where we saw most customers environments in the world. And we saw how the attacks starts to accelerate in velocity and increase in sophistication as the cost to perform a sophisticated attack is when going down.

That was about a year ago. it only increased much faster than anyone expected. And there is also a very large report at Crowd is published in March 26 on how the time to perform an attack went down dramatically. And another report. It's on tropic release in November 25 that basically saw the same thing.

When we think three years out in the future. We believe that most work will be done by agents. AI will direct the intent, sorry. Humans will direct the intent, but AI will execute it. And that we go for probably most, let's say, work in the world. But specifically for cybersecurity, we believe that adversaries and attackers will use AI to attack companies and they'll be able to do it much faster than previously possible.

And that means that on the defenders side, defenders need to adapt as well. And this is why we positioned Artemis as AI Native protection platform to help companies protect their environment, detect and stop attacks within seconds, and replace the traditional workflows that are very human heavy, and assume that attacks will take hours to days to weeks.

Dan: been waging an AI versus AI war my entire career. when we were at Twitter, it was various kinds of spam bots and various kinds of comment, bots and all types of different automated systems aiming to game the various Twitter algorithms and squeeze money out of people and scam people and juice, different kinds of ad placements.

And there's all kinds of machine learning models that we built to detect these kinds of automated activities and clamp down on them. when I was at abnormal, the, really, the first waves of AI in cybersecurity really beginning to be shocking in the scale of the change came when you had the ability to automatically generate emails that seemed extremely relevant and extremely, good English and very convincing via the early stages of generative AI is, it was relatively shortly after the first chat BT launch, that it was possible for attackers anywhere in the world to be able to generate extremely topically relevant.

Phishing emails and send them out in mass. We saw humongous amounts of that at abnormal. That was a humongous problem for many, customers that we built technology to defend against and protect against. We saw it increasingly get better and had to increasingly improve the systems that we built in order to fight against it.

And now, most recently, you have agents that are really a different paradigm than either of the automated bot attacks or the, AI generated texts, social engineering kinds of attacks that are structured to be able to perform these multi-stage reasoning and very fast exploitation. And fighting against this, in my mind, is somewhat of a natural extension.

I, I've just seen these techniques getting better and better, different kinds of techniques over time. it's very clear that this is how all the various adversarial games, that we have to play in order to keep people safe and build a safe, safe internet and safe world.

Josh: One of the challenges that a lot of AI startups are facing right now is that many of the enterprise customers that they wanna sell to aren't yet ready for the next gen thing they're building. So these companies need to build in legacy features for parody while also building to the future. do you tow this line or make that balance?

Shachar: So first, a, one of the, my favorite things about cybersecurity and the reason I spend my entire career in cybersecurity is that it's a very innovative field because you have to make a change. Like attackers will go after the financial gain and will use the latest technology.

And on the defender's side, you'll have to adapt. It's not like you should, like, you have to, otherwise you'll be compromised. And that drives a lot of willingness to adopt new things and change what we're seeing from customers. From even our customer base, about 70% of our customers replaced their legacy solutions with Artemis.

'cause they say, I'm clear that at this point in time I have to make the change right now in order to be prepared for what is already the reality. And the other 30% use us to augment their legacy themes and legacy solutions. 'cause they say, I'm clear that in two years from now I'll need to be completely off this legacy stuff.

I could just wanna cut it. But it's a long journey and there thereby I'm starting today in order to still have complimentary, let's say, enhancement to my overall strategy, but not have to reef and replace and change things too dramatically.

Josh: Why is AI native so much more important than AI enabled every. Legacy player is now scrambling to integrate AI into their platform. Can you give just a high level description in your mind of why a company that was built, AI natively has a massive benefit.

Dan: I think it comes down to the core control layer that you have built into the very foundations of the system. When you have, every kind of decision that needs to flow through traditional logical systems as the, foundational rails, then layering a on top AI on top of that is bottlenecked. It's harder for different pieces of the AI to talk to each other.

It's at some point things need to be compressed into this old language, this old way of different parts of the system talking to each other. It's much harder for this, it, what that ends up looking like when you have a AI enabled capacity on, on top of older features is you have a core foundation that's relatively unchanged, then a bunch of different product features.

And each of those product features have their own a new AI piece that's added to it. And if there's ways that those new AI pieces talk to each other, they either talk to each other through the mechanisms of the legacy system or they talk to each other through layers that are, that are layered on top, uh, between each of these different components.

And neither of those are, are the same in terms of the ability to actually have good performance and, and the ability to really represent something that involves learning an organization, learning the core things that drive decision making as something that's replaces those underlying guardrails, those underlying foundations. with that kind of legacy, log logical structure with something that is inherently agentic, that really takes this kind of reasoning layer and builds it into the very core of how the system is structured and how the foundations are built and how information is represented.

Shachar: And they're also stuck with legacy organizational structures, which is much harder to change. Like at least as hard to change as changing your legacy technology stack

Josh: You're both first time founders. What's been the most surprising part of stepping into the founder role for the first time?

Dan: I remember when I'd first thought about this, I was thinking the, what's gonna be really strange to me is that phase between zero and 20. Because the first company I worked at, I was employee number 15 or so, and, when I was at Abnormal, I was around number 200 and I joined Twitter, and Twitter was several thousand people.

And so I'd had experience in the, 20 to 80 range and experience in the hundreds range experience in the thousands range. But the, what is that zero to 20 going to look like? I wasn't totally sure at the beginning, but I was surprised by how natural it felt, how it was actually this.

This feels not that different. And from a, building perspective and relationship perspective, it's the really what has been different is the thing that I thought was going to be different, which is the level of ownership and commitment to level of, of certainty that this is the, this is the right thing for me.

I feel like I was built for this and I, I'd always felt like I was built for this and that has not been a, that has been the pleasant, I am pleasantly happy to see that it is as I expected. what was the surprise was how the zero to 20 phase was not quite as different or wild as I'd originally expected it to be.

Shachar: I was also presently surprised having the best time of my life, I'm living my dream and it's been a blast. Honestly. A lots of people are like, oh. It's really hard to be a founder and it's probably the hardest thing, especially professionally that I've done in my career. But it's also been by far the most rewarding thing that I've done.

And it's just like a lot of fun. There's like, yeah, I can't wait to sell the company. I can't wait to just like be in like an easier job. And I'm like, yeah, I really, I wouldn't trade it away. I've just been having fun and building with an amazing team, building for amazing customers. So for me, the, I guess the surprise was that while there are hard parts, nothing really seems hard, but more so like a fun challenge to crack.

Dan: And so we've done a lot of a hundred hour weeks in a row. I've, I don't think I've ever done this many, a hundred hour weeks in a row, but.

Shachar: I don't know if we had, we have done stop hundred hours.

Dan: but it is incredibly fun and incredibly wonderful, and that

Shachar: Yeah.

Dan: think both of those is very much as I expected.

Shachar: Yeah. And then maybe one, one accommodation for people at home. If it's really the thing, being a founder, building a company is what you want to do, like go do it. And there is potentially a time in the career that it's better to do it. Like both.

Josh: And what time is.

Shachar: I think it's personal trust. Both Dan and myself have waited relatively, relatively long in our careers.

Like both us were in senior leadership roles. We have both been an employee employees in startups that were acquired in massive acquisitions. We were early in this, employer, in these companies. so for me it was about. Seeing the experiences from each stage of a company and then incorporating that into Artemis.

Some people can also do it right after college and they'll still, be able to nail it. But I think it's a personal choice of like how you think about your career, but when you feel like you're ready, I would say just do it.

Josh: Are there any things that you are both personally thinking about as to how you hope or plan to level up as a founder,

Dan: I mean there's so many different areas where I personally want to improve. I want to be able to learn from, the experiences that we have, learn from each other, learn from the team. I think that there's a lot of things that's I'm doing for the first time. I've never been in a sales role before, to this capacity.

I've been in management roles, but it's different experience to manage people who have very different skillset sets, different kinds of, expectations and roles. And I think that leveling up as a leader in general, leveling up as a builder in the capacity of company building. I think are all areas where I want to continue to grow and continue to learn and continue to elevate.

Shachar: I think a lot about, being. A leader in a native company, which is really interesting because there is no playbook for it. There is no guideline on like what makes an truly a native company successful. And now we're starting to build in the next year, in 2026, we'll start building our executive team and we'll grow.

Right now the organization is still fully flat for like a very purposeful choice to allow people to just move fast and ship. And a lot of what I'm thinking over the next few years is that work is going to change how companies are built and operate is going to change. And I look forwards to growing with that broader, let's say, wave in the market.

And let's say position artes is the leading standards in how AI native companies operate successfully.

Josh: So I wanna wrap up with, my partner, Brett always likes to end the conversation by asking. Who has had the biggest impact of each of your careers and philosophy as founders?

Dan: Yeah, I would probably say my brother. he recently started a company not too long ago and, became very successful, very quickly, and, watching,his decision making process to leave a job where he was incredibly successful at, to start something. And then, the rapid success that he had, very early on, I think, and seeing the ways that he made these decisions, it really, I would say, opened my eyes to how much is possible when you see an opportunity and take advantage of it, and really set extremely high expectations.

Shachar: For me, uh, it's probably the founders of De Miso, the startup I worked for. I joined there very early and I joined specifically there because Deto was already their referred company, so I joined in the very first. I I told them that I'm Learn build a great company. And they gave me a lot of visibility and they also said, we'll give you the visibility and we'll give you the first check when you end up starting the company.

And they ended up really investing in Artese. They gave the first check, I guess you guys gave the first check.

Josh: Woo.

Shachar: it was all together. Money was in the bank the same day. But,I learned a ton there and a lot of how Artis is built this model on de Misto or especially around customer obsession, people being very strong generalist and people just owning things and delivering to customers and, they've been very meaningful and still help a lot in just thinking through how to continue to succeed and scale the company.

Josh: thank you both very much. Wish you the best of luck with the upcoming launch and continued success in building a kick ass business.

Shachar: Thank so much, Josh. Really appreciate you having us. It was a blast.

Josh: It was fun.