From Military Intelligence to Tech Entrepreneurship, Ariel Assaraf shares his insights into deep tech, customer satisfaction and building a unicorn

Robin: Can you share your journey leading up to becoming the founder of Coralogix?

Ariel Assaraf: Sure, my career path actually began in the Army. I spent some time in the intelligence unit where I learned the basics of data analysis and scale. This laid the foundation for my tech career. Later, I joined Fairing Systems, which eventually became Good Night, a NASDAQ-traded Homeland Security Company. In 2015, we founded Coralogix with three co-founders, aiming to modernize monitoring and observability practices.

Starting in the Army must have given you valuable insights into data analysis. What were some key realizations you had about data during your time in the military?

Being in the Army exposed me to the critical importance of data. It goes beyond just business decisions; it influences decisions in Homeland Security too. Learning to understand the impact proper data analysis can have on lives kind of got me into that world and sparked my interest in tackling big data problems, particularly in the realm of observability.

What specific challenges did you encounter that drove you to found Coralogix?

The ultimate challenge we identified was the duality of data analysis, which became the core of Coralogix. On the one hand, there's the need for real-time, high-frequency, and high-concurrency data processing to handle streams of data in high-frequency and high-concurrency that provides instant alerts. On the other hand, there's the requirement for ad hoc, long-term, and complex querying to uncover unknown unknowns. Balancing these needs became the cornerstone of Coralogix, especially in domains like Homeland Security and observability.

So the know knowns and the known unknowns are typically things that you want to answer in real-time. So if we take it to Homeland Security or observability or anything else, there are things that I know that are wrong that are not good. And if I know that they're wrong, I want to be notified immediately when they happen. There's no point in waiting. There's no point in running deep analysis. I just need to actually verify those things that I know every other second and make sure they're not happening. But then what happens when there's an unknown, when there's something that I hadn't anticipated? It is either already happening, I don't have an explanation for, or something that I'm trying to understand based on the data points that I have.

And that is typically a long-term and very complex query. Now the problem that we were seeing in and out and every market in any data market, by the way, is that those two worlds don't combine. It's two very different ways to compare them to a complete different example, but you can sell a Michelin star restaurant together with fast food at the same place; both are important. But if you try to prepare a burger for two and a half hours with a burner, no one's gonna buy from you. And if you serve at a Michelin star cost and serve me burgers and fries, no one's gonna come back. You got to figure out a way. How do you separate that?

But what I like about this example is that a lot of chefs here, and also I think globally, if you look at Gordon Ramsay and others, they have separated into two, right? They have their gourmet shop, and they have their fast-food kind of off-the-counter food because you can't have both living together. But in data, most platforms today try to combine both. And then what happens is that systems have had to be high memory, high disk to do those long-term queries, and on the other hand, they have to be high bandwidth to do the real-time stuff. Now it's possible, but you inflate the infrastructure to be something huge. And then what happens is? The limits of memory you need this you need and CPU. You need one to push the other. So it's like a three-dimensional issue; you need more of this for the long term. You need more memory for that. You need a high CPU for the throughput, which requires machines that come with more memory, and then one pushes the other and takes you to this massive system, which is a single point of failure, extremely expensive, and always has inherent latency. And that's sort of the problem that everyone was trying to solve and kind of stuff.

The thing is that until then, database companies tried to solve that by optimizing so I can run more complex queries. I can use less memory or less CPU to do the same. All are great incremental changes, and we were trying to figure out what would be a complete different architecture for this.

And is that what you've done with Coralogix?

We actually decided to do both things in a separate way in our pipeline. So what we decided is, instead of trying to store and then find meaning, there's a journey to that. Let me give you that journey until we got there. When we founded the company, that wasn't the idea. We knew there were a ton of existing tools. We knew there wasn’t a place in the market for another one. But we thought we could create a kind of machine learning layer on top of existing tools, simple, intuitive, slick. People would just plug it into their systems and run, get more practice. But that didn't work at all for various reasons.

One, it's not easy to integrate with these other products; two, they always have something that's not as good. It's not as slick, but it solves a problem. Three, the main one is we started from a problem of inflated infrastructure and cost. You're coming to me and saying hey, let's add more costs and more infra, and you're saying you're not even touching my problem. Right? It's like if I go back to my bad example of restaurants. It's like you're at this Michelin star restaurant getting a burger and fries, and I come in and offer to supersize it. It's supersized, but this is not my problem right now. And so for the first three years, we barely, and "barely" is an understatement, landed any significant customers.

Only in 2018 did we kind of flip everything. We really built out this architecture that analyzes the data in the stream, and then encourage the data from remote. That's that separation. We created two engines: Cold Streamer, Streama, is a stream analysis engine. It analyzes those high-frequency, high-concurrency data questions, known knowns and known unknowns. Dream only using CPU network almost being used. And then Data Prime, which only does long-term complex queries on the customer's own archive, which means low cost, long retention, and data ownership. And mainly uses memory.

I don't have to sustain that level of memory resources throughout the entire time. I only spin them up when a user runs one complex square, and in the world of Cloud native, I can spin up and down. So if you're a user that runs manual queries, a thousand, two thousand per day, the big amount of queries comes from the alerts. They run every minute, hundreds of queries every minute, but they're in stream. Then if a query takes 10 seconds, then I have to run a thousand queries for 10 seconds, that's 10,000 queries. That's a lot on one hand. But 10,000 seconds is just a hundred and fifty minutes.

It's just two and a half hours a day instead of having that same setup running 24 hours a day. So think of the cost savings. But now also add to that, everyone understands the cost and resource constraints. So they're not giving you the best performance possible. We can actually spin up big time for 10 seconds down. You have to keep that big time throughout the day. It will be disastrous. So that's things that we've done and this is where the company started growing fast.

And was there a moment when you felt like giving up, when you doubted what you were doing, and what did you take away from that experience?

I think at some point in the journey, we had to ask ourselves: where's this going? Is it going somewhere? Are we able to pull out something here? And actually do something meaningful? Is this market something for us? Are we the team for the market? Because at the end of the day, it's either a problem of the team or the market or the product. There's just three things people make, startups are very complex, but you gotta have the right team product and market.

At that point in time of change, we actually replaced all three. So I became CEO at that time. I'm one of the co-founders, and while I was leading product before, it wasn’t like I didn’t have responsibility over whatever other decisions we made before that. But at that point in time, it was evident that we needed more product focus on how we're leading this and learning the market. We changed the product. We revamped the entire thing. It was streaming. It wasn't batch. We were analyzing before storing and not the other way around. And then, we changed the market. So instead of attaching ourselves at the Enterprise level to big other products, we're now selling a full solution of replacement, with the benefit of better performance, better support, and better cost.

So we changed all three. It's almost like starting a new company because we remained in the same market. But it was somewhat of a pivot we made back then. I believe if we had continued on the same path, there would have been no real chance for survival.

And what would you say is the biggest strength is, coming from a product background to becoming a CEO?

I would say that in tech companies, and particularly in deep tech companies, CEOs have to understand both the pain points and the market. Take Datadog, for example, a major player in our space that we're competing with. We believe their CEO faces challenges with their business model, and many customers are dissatisfied with their costs. Consequently, we're witnessing a significant number of customers transitioning to Coralogix.

But I will say that as a management team, as a company, they made very good decisions along the way both product and market and messaging. They made very good decisions. Their CEO is actually an engineer by heart who really understands the pain points. Their CTO before that knew in and out of the product. There's no real way to sell in our world. It's not package it and I'll deliver; customers are very smart and advanced, competitions fears and you have to really understand what you're selling and selling on an actual pain point. A lot of the time I'll meet a customer and say I don't see the problem. Like if you're good, that's fine. And if you're trying to sell where pain doesn't exist, it's very difficult to scale that way and retain customers even more so.

Let’s talk about customer success, how does Coralogix make sure that customers are happy in the long run?

That's one of the biggest challenges in this space because success and customer success are complex. It's not easy to ensure customer success, and it's not easy to find the right people for it. We came out very early with a model of 24/7 support. We started by committing to under 10 minutes and under five today. We commit to under two minutes human response. In-app chat for our customers.

That's the first layer that answers 24/7. Above that, there are the TAMs (technical account managers). They have a joint Slack Channel and team channel with the customer where the important but not urgent things happen. New data sources initiatives, etc. And then on top of that, there are the account managers that overlook the accounts from a commercial perspective.

So that layer gives us the ability to ensure that our customers are getting proper treatment and are successful on the platform, and also our involvement is management. We are very involved in making sure that our customers succeed. I meet customers in person, whether it's in Europe or in India. I go and meet the customers, I learn from it a lot, and I think it gives me a really good perspective of the market and our product. Obviously, nothing's ever perfect, and so being there and kind of feeling what your teams are feeling is something critical.

And are those your three biggest markets, US, Europe, and India?

I'd say Europe, the USA and India have proven to be phenomenal markets for us. Initially, people advised against venturing into these regions, due to their complexity and cost sensitivity. However, India has emerged as a brilliant market with exceptionally talented entrepreneurs and engineers, and the CTOs in India are fantastic. Additionally, Europe is experiencing a notable awakening, particularly in the UK and Germany, but also across the rest of the continent. We're seeing a lot more modern companies, a lot of movement to the cloud. I think the open borders in Europe give it a nice boost.

You'd see most of the engineers working for companies; they're not from the country of the headquarters of that company. It's always you'd see people from Portugal, from Ukraine, from Hungary, from Czech Republic moving around, obviously UK, France, Germany, no need to mention, and it creates a really good kind of a way for them to thrive and enrich each other and creates really interesting products. And obviously, the US will always be the largest market for software in the world. So we're focusing over there, and we have actual headquarters. So we have one in Germany. We have a German company, we have an Indian company, and we have an American company.

Can you tell us about expansion and using cloud services? What are the biggest challenges with using, for example, AWS and some of these other big cloud services when trying to expand internationally?

I think the biggest challenge would be cost and control of cost. There are a lot of companies starting around that space to kind of give you better cost containment and understanding of where you are and how much you're spending. I think that's critical. Beyond that, I would say that without the cloud, half the products we're using today wouldn't exist. I don't think Coralogix could have existed without the cloud just because of the needs we have spread over 10 different regions. Scaling up and down with unpredictable traffic, there's almost no way to do that in a data center. So we're very happy with the services that we're getting from the cloud providers.

Let's go back to fundraising for Coralogix. How did that work? Where did you go for funding?

So, initially, when we started, it was local in Israel, which has a ton of small VCs. I think it's a really bubbling VC scene here. We did our seed, A, and B rounds here. Then we went outside. We did rounds with the Greenfield teams, an Israeli-American team actually, and then we added Brighton Park Capital as an international partner, which was fantastic for us.

At the beginning, it was very hard to raise the first couple of rounds— it was a crowded market, and we were first-timers with no actual proofs. And then, I remember the investor that made it told me that it was gonna be the hardest round that you've made from now. It's gonna get easier.

At that point, it's mostly numbers. If you execute and you have the right numbers, the right growth, and the right product, it's much easier to get funding. You're not trying to sell a dream. You're not trying to sell a vision of the future, concretely. Obviously, if you don't have the numbers at some point, you have to raise on numbers, and then selling the dream is not gonna make the cut. But for us, I think the BCD rounds were much easier than the previous ones.

What’s the startup scene like in Israel. You said it's bubbling right now. Would you say that it's bigger than it was when you started?

Yeah, when we started, it was the kind of startup nation, a lot of small-scale startups, a bunch of small funding, small VCs. Today, there are tens of IPO companies in Israel. I'm not even talking unicorns. I think there's over a hundred. Just to give perspective to people, Tel Aviv has less than half a million people. It's a very small city, and it has, I think, 2,000 funded startups in it. Just to give perspective. We're talking about one funded startup per 200 people or so. So you can imagine finding employees is not easy, but besides that, I think it's a really good scene, and people help each other, and the more mature companies try to help the younger ones, kind of being their beta or their design partners. I think VCs here are, because of the success and the IPOs and the exits, just this quarter, there were exits over three billion dollars in Israel with the ongoing situation and everything. So imagine, VCs are making a really good run here. And so that attracts more money and fuels up the industry.

Would you say that, as an Israeli, you have a specific entrepreneurial spirit or mindset, or do you think it's just the environment right now in Tel Aviv that's very conducive?

I think there are several decisions that were made here. Historically, Israel was one of the first countries where the government directly invested in startups, which was a very good decision. 35 years ago, it started. I think the whole absorption of the USSR after it broke down at the beginning of the '90s, a huge amount, over a million Jews came from there, and a huge amount of engineers came to Israel, looking for something to do. The country funded them, and they started the first startups, mainly semiconductors, and that exploded. Then, Galileo companies like Mellanox. We're talking billions of dollars.

And now, those people have become angels, and the small community makes it so everyone knows everyone. It's very easy to reach out to people, and it's a very friendly atmosphere here. Also, I think the nature of being in a stressful country, it's not like Europe; it's a lot more intense.Both from external threats and the cost of living in Tel Aviv, the most expensive city in the world, all these things combined create a very entrepreneurial atmosphere here, and a lot of people want to try their luck and start something.

Combining that with the talent and the training that you're getting in the Army, many people start their career and learn in the Army when they're very young and can still be kind of designed at 18 for work ethics knowledge, and so on. All those combined create a really good scene here.

Let's talk a bit about you and your background. Is your family entrepreneurial? Where did your idea to become part of the startup scene and to become a founder come from?

I would say yes, it came from home. But my biggest push, I think, was actually in the Army, where you see so many people that start their own companies. There's something about knowing someone and seeing him starting something successfully that kind of gives you the feeling that it's doable. It's not anything that's out of your ordinary. You don't have to be a superhuman. It's just about, obviously, a good idea, a good team, and a lot of hard work. We're talking about a company started over nine years ago.

But at least it makes it a lot more real, at least it made it for me and kind of pushed me into this. Then when you finish the Army at 21 and you get a really good job at a good tech company, it gives you a little bit of a cushion to actually try and attempt. So I think that was great, and the scene here again in Israel. When we just started, first-timers, no funding, no customers, no product yet, we were already applying and getting into three accelerators. Experienced tech people giving you one-on-one sessions, free offices, cloud credits, and so on—those things really make it much easier to start your business.

So for yourself, what do you want to do more of, say, for the next five years? What's driving you?

At the end of the day, I want Coralogix to become part of that wave of Israeli companies maturing and taking our country from a startup nation to a scale-up nation to a public good enterprise nation. If we take this company public in a couple of years, that will be great and create a lot of wealth here and for our investors, obviously, but also do something that changes a bit of this industry.

We've observed that Coralogix saves companies 70% of their costs. For me, it's not just about the financial aspect. Many times, choosing Coralogix over other products prevents layoffs in struggling companies. For instance, if we save half a million dollars for a medium-sized company, that means retaining jobs for several individuals, which is incredibly rewarding. Scaling this impact is something I'm passionate about.

And what about you personally, in your life outside of Coralogix, what are you running towards?

I don’t have much outside of Coralogix. There are personal interests, but obviously, more quality time with family is great. But then those things, how do you balance it all?

How is AI going to change the current landscape? And what do you think are the biggest things people need to be aware of?

Yeah, I think in short, the nature of questions we're asking our data is going to change dramatically. These questions are going to be more directed towards the business and the actual systems. So, we're going to transition from engineering intelligence to actual business-level intelligence and business workflows. Instead of asking which one of my services is slow or having errors, I would ask what services are causing me to lose the most revenue. The answer that I get will be tailored to the interface I'm using. For instance, from a marketing perspective, it'll inform me about areas losing money attached to underperforming campaigns.

Conversely, if I inquire from an observability perspective, I'll learn that a certain area loses money due to high CPU, leading to delays in customer interactions and subsequent churn. Essentially, this evolution will provide diverse paths leading to the same core issue, whether it's driving top-line or bottom-line growth for the business. As we leverage AI for this, it becomes the focal point of our efforts, and we have numerous plans in progress, working on several intriguing projects.

And for the very last question, with this development and security being absolutely top of mind, especially data security, how is Coralogix ensuring that as you expand, data security remains a top priority?

Well, first of all, data security is critical given our backgrounds and where we come from. It's part of our core. We ensure this through multiple means. We monitor ourselves rigorously, and external auditors regularly assess us. Currently, we are SOC 2 Type 2 compliant, ISO compliant, PCI HIPAA audited externally, GDPR ready, Privacy Shield ready, and now also FedRAMP ready, allowing us to serve federal agencies in the US. We're also Nimbus ready for other federal government agencies. With these measures, we regulate ourselves and undergo regular audits to stay aligned with the latest security standards and best practices.

Thank you Ariel.

Thank you very much for hosting.