Hiring While Hyperscaling with Ashwin Rajeeva, Co-Founder & CTO of Acceldata

(Joel Beasley at 00:00:02) Hello, my friends. Today we're talking to Ashwin, CTO and co-founder of Acceldata.io, and we discuss how to attract top talent and make their work meaningful. Challenges Ashwin is... Here we go.

(Joel Beasley at 00:00:33) This is the Modern CTO Podcast.

(Joel Beasley at 00:00:44) So yeah, tell me a little bit about yourself. How did you first get interested in technology?

(Ashwin at 00:00:49) Oh, yeah. So, let me go chronologically. I'm Ashwin. I got trained as an electronics engineer, actually, and then realized I was not perfectly good at it, right? And as a young graduate, had some interest in technology, I would say, especially computers and programming, but nothing really passionate. Not really passionate about it. So I wanted to do a little bit more theater and acting, maybe, you know, things like that at a point in time.

(Joel Beasley at 00:01:27) That's so good.

(Ashwin at 00:01:28) I don't know. Most of the people, you know... But I never thought, you know, I would be a programmer, right? So I did my engineering, tried a couple of things here and there. And at that point in time, actually, I graduated somewhere 2005, 2006. So kind of started talking to friends and, you know, programming a little bit, and then got involved in it. Went and did my master's in IT, and started working at a company called Thoughtworks, right? Thoughtworks is this global consulting company, right? And I did a bunch of things there. So learned a lot, especially early days of programming, learning from, you know, working for different accounts and enterprises as a global consulting company of sorts.

(Ashwin at 00:02:16) And about 2012, 2011, 2012, I said, "Hey, I want to, you know, go do something on my own." So with a couple of friends from university, I started our company, which at that point in time was into social media monitoring, right? That was a big thing, 2011, 2012. So we had a lot of companies which come in, and essentially Twitter used to have a fire hose, and you could listen to that fire hose. And then you could run things like sentiment analysis, and then you go to different brands and you tell them, "Hey, you know what? If somebody says that I got Coca-Cola and it was not fizzy enough, I can let you know," right? And we tried that for about a year, year and a half. And, you know, what I realized is that probably two things. One, good friends don't necessarily make good business partners, right? Good colleagues. And the second thing is you've got to be, you know, not just the timing of it or not just the talent of it, but you also got to look at, you know, the market, right? So at that point in time, a lot of this forward-looking social media monitoring kind of market actually existed just in the US, right? There was not much demand for it, you know, in the market, in India, especially in Bangalore. So we kind of shut it down in about a year or so. I started again. You know, I was pretty sure that I don't want to really go back to a job.

(Ashwin at 00:03:35) So I moved to Europe. I spent about five years in Berlin. I used to live in the middle of Berlin, and I ran a consulting company through which I worked with multiple companies, mostly as an architect, right? So I worked in ecommerce. I worked in banking. And through that, I kind of got back with some of my old friends and started working again at Hortonworks as an architect, just where I, you know, started working with my current co-founders again. We did a lot of work on big data systems, you know, making it more reliable, how to connect hybrid cloud, which was kind of trying to come up at that point, right? And at that point, we kind of realized that there's a lot of companies who invest in data technologies, you know, building data pipelines, investing in technologies like Hadoop. But the problem still is one of operating it, right? It's either a talent shortage or a tool shortage, or they don't really know. They bought this system and they don't know how to operate it.

(Ashwin at 00:04:35) So we kind of came up with this idea that, "Hey, you know, we kind of do what Datadog did for microservices, New Relic did for microservices. We do the same for data pipelines, right, from an observability, monitoring, operability perspective." Started the company in 2018, and it's been three and a half years now. You know, we've been fortunate enough that, number one, we got some really good people join us early on, and that helped us in getting some early customers here. And then our early customers were really well-known enterprises, right? So first customer was GE, like four months after starting up. And then Walmart, you know, DBS Bank, telecom companies in Asia. So we've been looking to work with kind of the enterprise-level companies with our observability platform that we built early on. And now kind of, you know, evolving into, with the current growth and investment, kind of expanding into more observability across clouds, especially for modern data stack, you know, systems such as Snowflake, Databricks, that ecosystem, right? So that's been kind of the journey from then to now.

(Ashwin at 00:05:47) And, you know, through the years, I've kind of evolved from largely being a programmer, then being an architect/consultant, and into now CTO. So this is kind of first for me, you know. I managed a big team. It was largely, like, solo, one or two people, right? And now we are like 140 people. So it's been quite a trip for me. Yeah.

(Joel Beasley at 00:06:14) That's really cool. So what was... I mean, it's, let's see, you said it's only been like three years, and you went from zero to 140 people. What has been, like, a big challenge for you in the scaling of that growth?

(Ashwin at 00:06:31) That's a great question. I think everybody would agree two or three things. One is the technology space itself, right? Depending on what you're good at through your career, right, you would probably choose a space in which to operate in. And out of all the possible spaces you could do a startup in, enterprise technology and data technology is generally hard, right? And not just in kind of finding a good product-market fit, you know. That means talking to customers, talking to enterprises, talking to really large—some of them slow-moving, some of them fast-moving—enterprises, and figuring out what would make sense as a product, as a service, as an offering to them. And the second is trying to find, you know, people who would join this mission with you, right? And in the current market, especially, you know, COVID and post-COVID recovery and the huge influx of kind of capital into startups, there's been like a huge talent war. And it's now kind of, who can hire the best will win, right, in general, in startups. So you see this just across almost all sectors, all types of startups, you know, whether they're doing data, you know, they're doing something in SaaS, or they're doing Web3. I think this problem all of us face right now is the shortage of talent and having people join you to build something, especially in the long term, because these things take time, right? So like you said, it's been three years.

(Ashwin at 00:08:06) You know, we have a view that anything meaningful would take a little bit of time to build, to verify, get value for customers, right? And having people come in, buy into this mission, and stay with you for that time, and, you know, stay excited and hopefully not leave you because they're bored, has been, I think, the single largest challenge. The tech is hard, but I think one of the learnings I've had is, you know, being a CTO is as much of a people problem as much as, you know, just pure technology sort of problem.

(Joel Beasley at 00:08:41) Absolutely. So how do you attract the top talent? What's what's working for you in that space right now?

(Ashwin at 00:08:49) Yeah. So we tried a bunch of different strategies. So the first few people who joined us were, you know, colleagues we knew from a long time. Almost friends, right? So we've been in the industry for a while, made friends along the way. People who worked with us in Hortonworks and other companies who kind of understand data and in general, a bunch of people who can work well with each other. And then, you know, post some of the seed financing and Series A, what we tried to do is to kind of, you know, try everything, right? Throw the kitchen sink at it. Try to attract people on LinkedIn, you know, get those licenses, ask for agencies to, you know, find developers, even try to hire in different geographies, right? And some of that was successful. But I think what I've learned is most of the time, it's not so much about, you know, what compensation are you giving, you know, what are your stipends and things like that. If you look at it from a candidate's perspective and a talented programmer, right, she or he is essentially looking for, in some sense, meaning, right, in their day-to-day.

(Ashwin at 00:10:02) And I always joke around that the number of programmers who are, you know, highly paid, most of the time at a desk job, free lunch, excellent offices, who just complain about their jobs—it's so, you know, frequent in our industry. This disillusionment, even in the best of the companies. And that's because I think a lot of people, when they start working in the tech industry, are also looking for some meaning, right, in work, right? So what do you do on a day-to-day basis, and how does it make a difference, right, in whatever context they are in? And I think what has worked for us is I translate the domain that we have been working on, right, which is this thing called data observability, which is, you know, a new term which we're trying to, you know, some practice which we're trying to bring to market. It's still quite a technical term, right? And it means a lot of things. But to translate that into meaning for a person who has probably 10 choices, right?

(Ashwin at 00:11:09) So I think about, hey, you know, why is data so important? How are enterprises looking at it? And why is this, you know, the opportunity, you know, probably of your lifetime, right, to create some value for yourself and for everybody—customers and all the stakeholders—versus another startup which says, "Hey, Web3, DAO, we are reinventing finance for the entire world," right? So for a talented programmer, you just have to find kind of the best way to fit meaning into the next three to four years of their lives. And more than anything else, I think that brings in some success. Now the caveat is that it's not easy to do it at scale, right? So you can't hire 50 people this way. And so even now, I mean, hiring is a challenge, and we get—we are really picky about who we get—but we make sure that kind of they are aligned to what we want to do, and they're in for the long term.

(Joel Beasley at 00:12:13) Yeah. That's super important because I mean, I know it's pretty common to kind of switch jobs every one to two years for engineers. And, like you said, with a long-term vision, that doesn't really work on your end if that's what they're doing. But so I'm not super familiar with data observability. Can you tell me, like, what is the problem that companies are having when they come to you for a solution?

(Ashwin at 00:12:41) Okay. So I'll explain it in a very simple manner. So observability itself is kind of a new term, right? And, you know, our industry kind of works on these—every once in two years, you've got to bring in a new way of looking at maybe an existing problem or extend something. And largely, the industry works like this. And observability is a kind of an evolution of systems monitoring, right? So you had Nagios and a bunch of other systems. And what we said is, "Hey, you know, everybody is going to now build systems on top of scalable infrastructure such as AWS, GCP, and others. And it's probably not enough that, you know, you know about what's going on in your system when there's an emergency, right?" So you get an alert saying, "Hey, your disk is now gone." It's already gone by the time you get it, right? So now you've got to have this whole operation where we'll wake people up who know about it, and then hopefully you fix it. And then your website shows up something like, "Hey, we are down for a while, and we'll be back." That's been the pattern for a long time. So observability is a way to kind of, you know, put enough probes into systems so that you just don't get informed of it when things go bad, but you essentially know the health of your system. You understand what's really going on, right?

(Ashwin at 00:14:03) So that if your disk starts misbehaving, you probably know because there's a big dashboard in your office which tells you what's the state at this point in time, not when something similar happens. And over a period of time, I think the way we build applications and mobile apps has been, you know, it's largely consolidated. You say, "Hey, you've got to do smaller services, individual responsibilities." We try using containers, and we send it into something like a Kubernetes or into the cloud in some deployment environment. And even the observability for these systems is kind of, you know, now it's a standard stack, right? So you have Prometheus, you have Grafana. You largely are monitoring your web sessions. You're monitoring your database connections, and you're looking at—you have all your events, metrics, logs, and, you know, traces, the telemetry data going in. And what a NOC engineer or a SiteOps engineer sees is a set of dashboards, which at any point in time can tell them the health of their application, right? And that's what New Relic and Datadog do, essentially.

(Ashwin at 00:15:24) If you look at in the last five to six years, there's been an explosion in data technology itself, right? So Snowflake, Databricks, Spark, even Hadoop, and friends, right? And, you know, Starburst and Dremio—there's a huge explosion in general availability of data analysis and people using data to make real-time applications. So if you go into a bank's website and you say, "Hey, I'll give you five minutes," the bank will give you, within five minutes, your eligibility for a loan or for insurance. And what they're doing is essentially, you know, looking into a data system, into a data lake, to churn a lot of data and then take this decision. Now, if these systems are so important to your business, yeah, because your business is now so data-driven, then it's important that you observe and monitor these systems just the way you observe and monitor your front ends and your mobile applications and your back-end services, right? But the knobs that you need to turn in data applications are entirely different than the knobs you need to turn on a, you know, microservice or a back-end application, right? You might not be so interested in, "Hey, what's my connection limit? What's my pool limit? What's my disk?" right? But you might be interested in, "Hey, did the data move correctly? Did it come on time? Is it the right data on which I can actually take action?" right?

(Ashwin at 00:16:45) And so since the knobs we need to turn are different on the data side, right, so what we wanted to do as a company, our hypothesis and our effort has been to build an observability system for the engineers who build, you know, data infrastructure. So data lakes, lakehouses, data warehouses, data pipelines, machine learning systems, and figure out two things. One, what is it that they want to observe, and how can we actually go ahead and build it, right? So our customers typically tend to be larger enterprises who have huge investment in data-related technologies, right? Whether it's machine learning, on the cloud, on-prem, doesn't matter. Have, you know, petabytes of data, terabytes and petabytes of data under management on which they run different workloads, and they want to understand the state of these systems. And we have kind of, over the last three years, custom built a lot of applications, connectors, probes, which can actually get this information into some sort of dashboard, which is actually—

(Joel Beasley at 00:17:54) Okay. So is it, like, making the data understandable to the rest of the infrastructure to be able to operate on it and turn the knobs? Am I understanding that correctly?

(Ashwin at 00:18:06) Yeah, that's in the ballpark. Right? So let's assume we wrote a microservice and you put some logging in, put some probes in. Every time it goes bad, you get an alert on your page.

(Ashwin at 00:18:19) It will be saying, hey, your request per second is going down. People are getting timeouts on your website. Right? As opposed to what we do on the data side, if you were using our software, it will tell you, hey.

(Ashwin at 00:18:30) You were expecting 10,000 rows to be loaded and a job to be run in about five minutes, but 5,000 rows loaded and your job didn't even run. Right? And that's going to impact downstream systems. So the outcomes are similar that you are monitoring a mission critical, business critical system, but the knobs or the meters you're looking at are pretty different between data systems and applications.

(Joel Beasley at 00:18:56) Okay, cool. So can you tell me about the three products, the Pulse, Torch, and Flow?

(Ashwin at 00:19:04) Yeah. Like I said, ours is a more enterprise focused platform. Right? So we typically have been working with larger companies. And the way a lot of, let's say, data engineering teams operate is that the data plane, where the data moves in through, is independent from the infrastructure on which you operate.

(Ashwin at 00:19:29) What I mean by that is you could take the same data, load it into a database. You could load it into a Snowflake, or you could pump it into a relational database. So it's something like a Snowflake. What we're saying is that, hey, I don't need to install anything. So it's cheap for me. I don't need to operate something, but I can pump in a few terabytes of data and run ad hoc queries. So it's cheaper than hiring an entire data engineering team to run it. So that's really what we want to do. And since most of the data pipelines are designed with this architecture where the infrastructure, the data, and orchestration is different, we thought the right way to attack this problem is to build a solution for each of the places. Right?

(Ashwin at 00:20:05) So Pulse is focused towards infrastructure and platform. Right? So it's as simple as saying that, hey, if you're running your workload on Spark, then we have the probes to monitor Spark. If you're running it on Snowflake, then we have the probes to monitor Snowflake. What does monitoring Spark mean? Hey, am I running it correctly? Is it parallel? All the recommendations that you need to make sure that when you write a Spark program and your infrastructure is tuned correctly, we make sure that we can observe all of them. And then you take the right decision whether you are getting what you paid for in terms of infrastructure at all.

(Ashwin at 00:20:57) The second bit is the data itself. Right? No infrastructure can work well unless you have the right data moving through it. Right? So let's say you loaded data from your database into Snowflake and all the last names are not. Right? You would never know unless you are also looking at the data. So as data moves between the systems, as you load data from system A to system B to system C, eventually for consumption, you've also got to look at how this data is doing. And so there are norms you can look at. For example, is the data valid? Am I comparing two tables together? Has my average age in my table shifted from 35 to 65, which is impacting my algorithms? Those aspects are kind of observed by this layer we call Torch.

(Ashwin at 00:21:33) And the last one is essentially an implementation of open tracing, sort of an open tracing implementation as applied to data technology. You can describe your system as processes acting on data. And what the Flow system will do is that it's going to give you like a command center. Right? It's going to give you, hey, your data moved from source A, source B was processed by this program into D, and from there it moved on. So it actually kind of tries to represent visually the path your data takes in your organization. It's a little bit more involved because the way you do it is through SDKs and APIs. Right? It's the most, in terms of implementation perspective, it's the most inclusive. But we believe that unless you do that, you have no idea what your engineers did over the last three years. Right? Everybody came in, bought some technology, pumped it all together. Now it's a big ball of wires. And unless you have that information, you really can't take much of an action.

(Ashwin at 00:22:44) So for us, data observability is all of this and not a data quality plus-plus or monitoring plus-plus. And so what we have done is even though it was riskier at the time to kind of work on three things in parallel, our belief in this is quite strong. So we said, okay, we're going to invest in building some of these. So out of that, most of our customers now use Pulse. We have good traction on Torch. We have a few Lighthouse customers. Some of these are big names, big enterprises. We are still kind of in the Lighthouse customer discussions and discovery phase of Flow.

(Joel Beasley at 00:23:33) Cool. So with having to API in for Flow and the level of involvement you have to get to for Flow to work properly, I imagine you have to have a lot of security around that, obviously dealing with enterprises' data. What does your security function look like? Do you have security as a function at Acceldata? What's your team structure around that?

(Ashwin at 00:24:03) Well, we're learning some of it as we go along. Right? So even though there's a lot of agreement between the tech folk on what good architecture should look like, once you kind of ask your customers for how their stack has been laid out, you realize that there's nothing common in between that. Right? Everybody is using AWS in their own way. There's some semblance of repeatability. But more often than not, you have custom pipes between data centers. Some of them connecting AWS on the cloud all the way to their data centers, all kinds of security restrictions. So it's hard to build for something like this.

(Ashwin at 00:24:48) So at this point in time, we don't really have a security function. We knew this. Right? So we come from an enterprise background, so we understand that this is going to be a problem. So as a part of building even the MVP and going into production, we have kind of been putting some of these practices in. Right? So for example, every enterprise that you go into would, or any company you go into would, request you for vulnerability scan and penetration test reports and things like that for your own software which you're going to put in. Right? So we have made sure that a lot of this is kind of available from day one because it's more or less a prerequisite at this point in time. Right? So you've got to have all of the information necessary even before you make a sale to some of these larger customers. Right? Otherwise, you're going to be shown the door immediately after the POC if you can't comply.

(Ashwin at 00:25:44) So it's more of a, instead of kind of retrofitting it later, we've been putting in security as a practice with the developers as a part of the product from day one. Of course, we can't do it for all types of customers, but at least we've seen 80 to 90% of security concerns we already have the answers to. And the ones we don't have, we kind of try to figure it out. So we have done everything from installs on data centers where we downloaded all the images and went to the—during COVID times actually drove to the data center, installed it. Right? We have installed it over the internet. During COVID, we did it over Zoom. And we have also done air gap installs at this point. So depending on which customer wants what, we've kind of built in those modes already.

(Joel Beasley at 00:26:41) That's really cool. So I wanted to ask you a little bit about R&D stuff, because I saw that Acceldata has several patents. And recently I actually got to interview this company called Cactus Communications, which they help scientists publish academic papers. Well, that's how they started. And then they also just spun out this huge R&D side of the business called Cactus Labs. And so that's just been on my mind. And when I saw that you guys had multiple patents, I was just curious, what kind of R&D investments are you making at Acceldata?

(Ashwin at 00:27:22) I think it depends. So if you look at an enterprise company or you can look at companies in this space, in general data management, a lot of companies which have been ultra successful—the Confluents, Databricks, something like Snowflake—essentially build on one of two things. Right? An existing, well-received piece of software. So if you look at Confluent, you're working off Kafka, which has been around since 2009. And then five years of it being in the wild, you kind of build a company on top of it. Right? Because now people have actually already adopted it. If you look at Databricks, it's the same story. Dremio is the same story. Starburst is on Presto. So a lot of successful companies are building on top of five to six years of open source R&D. Right?

(Ashwin at 00:28:09) Whereas we kind of started off from scratch. So there's no existing data observability open source solution which exists, which has been widely adopted and all the security patches have been fixed. A lot of very intelligent people from different companies like Uber—it's none of it. It's just starting from scratch. So one of the challenges when you do this is that you've got to be talking to customers. You've got to be building product that customers want, and you've got to also have a product to kind of go through the customer. Right? There's a catch when you do that. And so what we have been doing is trying to get people who have been at least in the data space for—and since we kind of work from a data background, the first 10 people who joined us from an engineering R&D perspective were people who have contributed to open source Apache products, people who've been involved in the Hadoop ecosystem, either as practitioners, committers, or even operators. Right? So a lot of our customer success team comes from a real solid data engineering, data science background. Right? So that's kind of helped us create the core.

(Ashwin at 00:29:31) And at this point in time, we've wanted to double the engineering team. So that's the kind of investment we're looking at. But like I said earlier, due to the domain of data management, you need to find people who are comfortable doing data engineering and also application programming. It's hard to get people in that space. But the investment, what we want to do over this year and the next, we want to at least 2x to 2.5x the current engineering.

(Joel Beasley at 00:30:03) Wow. Yeah, that's awesome. I mean, 2x to 2.5x is like double the amount of people you have there hiring?

(Ashwin at 00:30:11) Yes. So, you know, double or 3x the number of people we have in engineering.

(Joel Beasley at 00:30:16) Oh, awesome.

(Ashwin at 00:30:17) In terms of the ambition that we have. Right? So if you look at what we want to do, we have these three parallel tracks running under one platform umbrella. And working with really big customers means there's a lot of being required. Right? How do you actually solve problems? Just like you said. Right? How do you solve problems around security, deployment? You need different ways of looking at things, SDKs. So the spread is quite wide. And to kind of fulfill this ambition that we have, we kind of need to really 3x the team that we have. And that's been the single largest challenge for me as a CTO facing right now.

(Joel Beasley at 00:31:02) What's like one thing that you wish you knew before you started as CTO?

(Ashwin at 00:31:09) Yeah. I think there's a ton of things. Right? So I think this has been the most challenging thing I've done purely because I always thought that building a company, especially a deep tech company, or an enterprise tech company, is a function of getting a lot of smart people together and basically coding a lot. Right? Putting systems together and building cool stuff essentially. Right? People and going to customers saying, hey, we have this really cool piece of technology. Why don't you use it? But what I realized is that just because you think that something works for you doesn't mean it's going to work for the customer. And it's the same in hiring as well. Just because you think that this is the most amazing problem that you can work on, it doesn't mean that other people are going to really find value in what you are doing. Right?

(Ashwin at 00:32:06) So in some sense, I think what I've been trying to do in my role is to take myself out from the architect/programming mindset, which I've developed over so many years, and put myself in the shoes of other people who, without them, you can't really build a company. Right? So even things like, hey, how do you place your offering? Right? How do you approach the market? Right? How do you look at the solution that we have built from a marketing perspective? How do you look at it from a success perspective? How do you look at it from a customer's perspective? Right? So customers coming in and asking questions around, hey, tell me how it's the ROI. For me, it's obvious. Right? I've been in this field for so long. Things are kind of obvious. Hey, this is, of course, it's valuable. But taking it from a customer's perspective and also communicating the value of what we are building for them so that they find kind of a shared motive with us as the founding team, as the leadership team, and then kind of aligning them towards the next five years of building something valuable. Right?

(Ashwin at 00:33:03) So I think that change in mindset where you're going from an introverted programmer into more of a management role where you're able to empathize with how others look at you, your company, your placement, and their work itself, I think has been the biggest shift that I had.

(Joel Beasley at 00:33:49) So as you 2x and 3x your team, how do you make sure you scale the culture and the values that you deem really important to the business?

(Ashwin at 00:34:03) I think you've got to spend time with people as you scale. And this is something which has come about in the last few years because of the persistent need to do things at hyperscale. Right? Really build fast, break fast, fail fast. And then if it doesn't work, then try something else. And in some sense, I think it would work for a lot of people. But for me, I think you've got to do something for a long time before you start seeing results on it. Right? And that is not just about building a company, also about growing any sort of audience. Right? Whether it's for data observability, whether it's the Modern CTO podcast, the first few times that you do it, you are going to run into issues. So you've got to spend time and that's the same thing, the same principle you apply to people and to culture as well.

(Ashwin at 00:34:48) So what we did is I think, between when we were 10 or 15 people, we decided on very few set of things because we really don't know how far and how big it's going to go. And we put in two things. One is we want to get people who are responsible. So who can operate independently. So if I get a customer success manager, I expect them to know customer success and to act in the best interest of the company. And the second, I think, important thing, which we've tried to maintain as we have gone from four to six to eight to 140 now is that we want to be generally nice to each other. Right? So no matter what happens, no shouting, none of it.

(Ashwin at 00:35:50) And that right now, what we do is, you know, anybody who joins us for the first fifty, sixty hours, I've spent time, but we make sure at least one or two of the leadership team who's been around for a while get to spend time either as a mentor or a manager and spend some time in kind of putting in these values as a part of the onboarding or as a part of the mentoring session as they join. Where you essentially say, hey, you're empowered to act in the best interest of the company, and, you know, while kind of having a good time at work, but you're not entitled to being rude or, you know, being a total pain for other people to work with. And a lot of it has actually helped us during this pandemic time because we hired a lot of people whom I've never met. Right?

(Ashwin at 00:36:46) I've probably spoken on the call, and I've never met them in person. And so trying to get some time with everybody and try to communicate this has been critical. And that's worked till now. I'm not really sure, you know, if we become a thousand people, how is that gonna work? But, you know, what we want to do is hopefully bring in other people who know this better than me.

(Ashwin at 00:37:08) Right? So people who have done this before and can kind of scale this organization much more effectively while keeping culture impact intact.

(Joel Beasley at 00:37:19) Well, it sounds like you're at a super exciting time at the company growing as fast as you are. Hiring is a good problem to have, just trying to keep up with everything you want to do. So if there's data scientists listening or other people that are listening and excited about what you're talking about, how can they get in touch to talk to you about joining Acceldata?

(Ashwin at 00:37:45) Yeah. So, you know, these are really exciting times for the company. We have come to a place where I think the problem is not about, hey, you know, do two people or two customers find value in the product or the platform. I think we have passed that phase.

(Ashwin at 00:38:00) We have really good customers. And now the problem, you know, like we discussed is scaling. Right? So we gotta go from a startup now to a company which can get hundreds of customers. Right?

(Ashwin at 00:38:13) And to do that, we need to scale all of it together. Right? So we are hiring for engineering, customer success, data scientists, data engineers, outbound, inbound sales, marketing, all of it. And a lot of these you can find on our career site. So if you go to acceldata.io/careers, you know, you can apply at the website or you can, you know, there's also an email over there.

(Ashwin at 00:38:36) You can just write to us, [email protected], or you can write to me directly at [email protected]. Right? So and we'll be more than happy to have talented people coming to us.

(Joel Beasley at 00:38:49) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.