Aran Khannafounder and CEO of Archera
Jon PenlandCOO at Kinsta
There are many critical factors that contribute to the success of a company, and having a great leadership team is definitely high on this list.
But if you don’t take the time to carefully set out your organizational leadership principles, your business won’t have a guiding star.
In this episode of the Reverse Engineered podcast, our host Jon Penland welcomes Aran Khanna, the founder and CEO of Archera. They chat about the importance of having leadership principles, encouraging your people to take ownership, and why a customer-first mentality will put your company on the right track.
- Every organization needs to have its leadership principles. Well-implement leadership principles lead to positive outcomes, both in company operations and customer relationships. So, it is essential for every company to figure out their principles early on. Aran explains, “We started with having, of course, a set of leadership principles that we’ve been tweaking over time. These things are never set in stone, but our HR contact helped us hone that into a set of operational hiring practices, including questions and things like that that I think really helped us bring in a more formal process.”
- What is the customer-first leadership principle? One of the most common leadership principles is the customer-first mindset. Almost every successful company in the world focuses on putting customers first. Aran talks about this mentality. “We are a fiduciary for our customers so that every decision that we try and make is driven by customer requests. Most of our roadmap is customer requests, and then, especially because we’re automating and sharing risk with our customers, everything needs to be made in the best financial interest of the end customer. There’s no world in which we would screw over one customer at the behest of another, for example. We have to be a fiduciary for each and every customer that we service.”
- Extreme ownership is essential. Taking ownership is the key to a successful career. If you want to grow professionally, extreme ownership is necessary. Aran explains, “If the site is broken, for example, or data is showing up in a way that’s maybe not ideal, make it so that anyone can take ownership over that. There’s no guarding your lane. If something’s wrong for the customer, you own it if you’re the one who sees it, and you can go and act. Otherwise, escalate it to the appropriate places, but there’s no like, ‘Hey, look, I’m not going to act on something if I see something wrong,’ and internally, we try to empower people to that end and operationalize that in a number of ways throughout the company.”
Today’s Guest: Aran Khanna
Aran is an engineer and an AI scientist who has worked on various popular apps and open source projects, including Apache MXNet. He’s also worked for tech firms of all sizes, including AWS DeepLens, where he was a tech lead.
Automation requires good prediction
“The problem that I saw when I was at AWS was that we had these awesome vendors doing visibility and native tools for visibility, but customers still couldn’t accurately plan and predict. And because of that, they couldn’t automate. And with Google, for example, and with all the public cloud providers, you’re billed by the second for the resources you’re consuming. So your optimization happens when your developer gets to it. It’s a one-time thing — maybe once a quarter — and all the time that you’re not acting, you’re losing money. So without good prediction, which is the first piece of what we do, you can’t really automate effectively, particularly around things like commitments that don’t impact your underlying applications but can really heavily impact your spend.”
Working remotely makes sense for some companies
“There is a lot of value to coming and collaborating in person, but there’s no reason to go into an office if you don’t have a reason to be there — just to sit in the aura of other people; it doesn’t make sense to us. So we try to throw team events like happy hours, quarterly retreats, and things like that, to get people together. And people will come together and fly to an office — they flew to Austin recently to have a design jam with the team down there, especially now that things are opening up. But I don’t think there’s any world in which we all need to go to the same place and be restricted to a talent pool in one geographic region.”
Hiring the right people is key
“Obviously, as a CEO, the biggest thing is people. Do we have the right hiring roadmap? How are we thinking about attacking this? And is this aligned with what the customers and the partners that we want to work with are going to be demanding of us three months from now? Because you’ve got to have a plan for all that ahead and make sure you have the right leadership.”
Explore your possibilities
“What was helpful to me on the journey was — particularly talking about earlier stage people who are maybe in college or coming out of college — using those early years to get a really good breadth of what’s out there, maybe working at different companies. And honestly, you don’t have to work at a startup immediately after you come out of college. In fact, in many ways, it’s better to go and do your exploration through your internships.”
[00:00:00] Intro: This is Reverse Engineered.
[00:00:04] Jon Penland: Hey everyone. My name is Jon Penland and Reverse Engineered is brought to you by Kinsta, a Premium Managed Hosting provider. In today’s episode, I’m speaking with Aran Khanna, founder of Archera. Aran, welcome to Reverse Engineered.
[00:00:16] Aran Khanna: Hey, thank you so much for having me, Jon.
[00:00:17] Jon Penland: Yeah, we’re thrilled you’re here with us today and to get us started, can you introduce yourself to our audience?
[00:00:23] Aran Khanna: Yeah. Well, my name’s Aran Khanna, I’m the co-founder and CEO of a cloud, SaaS, ISV platform called Archera. Uh, we make cloud FinOps, uncomplicated, automatic, and low-risk for our customers. And uniquely we sell insurance against the cloud commitment plans, among other things.
[00:00:39] And I’m sure we’re gonna get into that. You know, before that I worked at AWS, I worked at Azure, unsurprisingly, and I was also a researcher in Deep Learning and Privacy and did some fun digital privacy projects that I’m sure we’re gonna talk about, as well.
[00:00:51] Jon Penland: Yeah, absolutely. So, I do wanna start off by talking about Archera, which is what you’re focused on today, recognizing that a lot of our listeners are not gonna be developers or engineers.
[00:01:02] Can you describe what it is that Archera does in layman’s terms?
[00:01:07] Aran Khanna: Yeah. You know, very simply, when I was at AWS, there were tons of resources, you know, like VMs and managed services, things like managed WordPress, for example, that customers were able to buy through the platform and they could commit to resources in different ways, save money in different ways, which was very complicated.
[00:01:23] And really what we’re trying to do is make that process really simple through automation and make it really low-risk for customers through insurance. So, we’re sure that, you know, even if our robot messes up, we have a contractual obligation that we will, you know, save you money, not leave you on the hook with anything.
[00:01:37] So, you know, that’s kind of, in simple terms, what we do for businesses make it way easier to save money and super low-risk to, to commit to resources, using our, you know, insurance that we uniquely develop for the public cloud today. So, Azure, AWS, Google.
[00:01:50] Jon Penland: Yeah. So, moving, I guess, one step closer to your actual customers. I think, that’s a great sort of layman’s high-level explanation. A lot of our listeners are gonna be a little bit more familiar. And so, I’m just gonna use Kinsta as an example here because we use Google cloud. Our entire platform is built on Google cloud. We don’t work with anybody to optimize those costs.
[00:02:12] We do have committed spend contracts. We’re familiar with all those different things. How does Archera work with a company, like Kinsta, to help us optimize what we’re spending in the cloud?
[00:02:24] Aran Khanna: Yeah. So, you know, there’s a lot of ways, just in Google, today that you can see what you’re spending…
[00:02:28] Aran Khanna: …and, you know, it’s pretty good at showing you what you spent yesterday.
[00:02:31] And even if you spend the time and invest in visibility, you can, you know, really get granular in terms of where the spend is coming from. The problem that I saw when I was at AWS was that we had these awesome vendors doing visibility and native tools for visibility. But, you know, customers still couldn’t accurately plan and predict.
[00:02:47] Jon Penland: Mm-hmm.
[00:02:47] Aran Khanna: And because of that, they couldn’t automate. And, you know, with Google, for example, and with all the public cloud providers, you’re billed by the second, for the resources you’re consuming. So, you’re, automate your, you know, optimization happens to be when your developer gets to it. It’s a one-time thing, maybe once a quarter, and all the time that you’re not acting, you’re losing money.
[00:03:04] So, without good prediction, which is the first piece of what we do, you can’t really automate effectively. Particularly around things like commitments that don’t impact your underlying applications, but can really heavily impact your spend, you know? And then, finally, without the confidence to commit to a really high level, you’re not gonna drive meaningful savings.
[00:03:22] And how do we get that confidence? It’s really through this insurance product that we provide. So, when we come into an organization like Kinsta, you know, we install with a pretty, least privileged, you know, restricted credential that just lets us see what’s going on in the bill. We’ll help you build a pretty good forecast with uncertainty and all that, of your future usage.
[00:03:39] Put an automation plan in front of you, the platform will, you know, do a lot of this automatically for you. It’s pretty self-service and you can see, “Hey, this is what the plan is, from Google today. This is the plan that Archera can give us, incorporating their forecasting model, what they think an optimal blend of commitments are, and the risks that I’m willing to put into this, given that they give me insurance and have a bunch of ways to de-risk that for me.”
[00:03:59] And you basically get an apples and, you know, apples to apples of what you’re saving today and what would the automation and all this fancy stuff, uh, you can click a button and save tomorrow. Yeah. So, it’s very simple. We try and make it an easy button for a small company like yours, to really get the leverage of having a full-time person working on this without that cost.
[00:04:17] Jon Penland: Right. Right. Well, that was gonna be my next question because we are, we don’t have a full-time person working on this, but we do have a couple of folks who routinely and by routinely, I think we’re down to daily, are logging in, looking at our spend, buying new commits and trying to really optimize how much we’re spending because the difference between, for those who aren’t buying directly in the cloud, the difference between committed spend and just on demand is, it can be like a factor of three plus, right?
[00:04:43] So, it’s huge, a huge difference in cost. So, I guess the potential selling point for a company like Kinsta, which I have a feeling is close to optimized, the other benefit there then would be time savings. ‘Cause right now we are building out our own billing dashboards with APIs to make sure we understand what’s going on.
[00:05:02] We have people logging in and buying commits all the time. And, and so, it sounds like maybe you’re offloading some of that workload into software.
[00:05:11] Aran Khanna: Exactly. So, you know, we kind of have three pillars that we think about, in the platform. You know, one is cost, attribution and reporting, budgets, forecasting.
[00:05:19] We essentially wanna, you know, give that away for free, to our customers. We think that, there’s a lot of companies that I saw at AWS that charged one to 3% of your bill to go do that for you across, you know, all your platforms. I thought that was a little user risk, like, this should be something that’s a really good version that, you know, helps you build good forecast of the future, could be free.
[00:05:35] You know, then, the automation piece, we wanna make that a flat, simple, easy rate, you know, no percentage models, which misalign incentives there. So, it’s pretty cheap. And then, really where we make the money is where we share risk with the customer. So, right now I’m assuming that there’s some workloads that you don’t wanna commit for all that time.
[00:05:50] If you have seasonality, for example, in your business, you, you know, that stuff, you just run on demand. Well, we have a product, for example, um, this isn’t on Google today, but we have it in AWS where we can say, “Look, commit to us for 30 days. And you can cover a hundred percent of that and have a capacity commitment against it.”
[00:06:05] And, you know, we have gaming customers, we have eCommerce customers, they have those busy months, right? And then, as soon as you don’t need it, after those 30 days, you can give it back to us. Hold it as long as you need to, maybe that spike is the starting of a new era for your business, where you just keep growing, who knows. Right?
[00:06:18] But it’s that additional flexibility, those new strategies that we unlock that really, for an advanced team like yours, would be the driving factor. And also, that’s where we think, in the limit, we wanna make all of our money by actually sharing risk with the customers, by putting skin in the game, instead of helping use the Amazon or Google or Azure tools better.
[00:06:36] Jon Penland: Yeah. Now, you’re right. There, most of our spend is pretty predictable, but we do have some spend that if it’s tied to, like, a single user or a single customer on our platform, it’s just too risky for us to commit to that spend on a three-year basis or even on an annual basis. Because our terms with our customers say, “Hey, you can leave any time and we’ll give you a refund of unused fees.”
[00:06:59] Right? And so, because we don’t have that commitment upfront from our customer, if some of that spend is tied just to a single customer, it’s just way too risky for us to commit without, you know, some sort of insurance that protects us against, against that risk. So, that makes a lot of sense. So, kind of looking at Archera, backing up a little bit, what does the company look like today?
[00:07:21] Where are you located? What’s the size of the team? Are you in an office? Are you remote? Yeah. What is the sort of shape of the company itself?
[00:07:30] Aran Khanna: Yeah. I mean, the company is about three years old. Uh, we raised our series A in the back half of last year, I think, around the beginning of October of last year is when we announced it.
[00:07:38] And, you know, we’re 20 people now, distributed across three major GOs, but, you know, it’s, it’s a remote-first company. When we raised our seed round, it was February of 2020. So, the first big hires we started to make were all over Zoom. You know, the business was essentially born over Zoom, in many ways.
[00:07:54] I’d rather grew over Zoom, kind of through its adolescence, which is where we are today. And, you know, the three major hubs are Vancouver, Seattle and Austin, Texas right now, you know, some folks just are nearby. We have people in Phoenix for example and they’re associated with our Austin hub. But, yeah, we operate completely remotely.
[00:08:15] And, you know, we have a number of large enterprise customers, as well as tons of smaller startups, maybe like Kinsta, potentially using us. Right. Largely geared to AWS, but we have kind of Azure in a public beta and we’re working on getting Google kind of to the point where we can offer all the, the great insurance stuff, like the 30-day commitments by the end of the year.
[00:08:34] Jon Penland: I happen to notice that you mentioned series A and you mentioned remote sort of in the same context. And I’m curious as you were going through that process of trying to say, “All right, we have this idea. We need some funding to see if we can move it to reality. And then, you start building out that team. Was that remote element a challenge through that process?
[00:08:56] Because there’s, there’s this sort of idea out in the space of startups that, that a lot of potential investors don’t necessarily want to invest in a remote-first company that they prefer that sort of old school office culture. Did you find that to be a challenge, at all?
[00:09:09] Aran Khanna: No, I don’t think so. And particularly in the infrastructure space, you know, especially with all these open-source products that become infrastructure companies, remote-first, I think, is built into a lot of the DNA of the investors and the investors we have are very familiar with the space they invest in, in infrastructure, dev tools, et cetera.
[00:09:26] So, I don’t think it raised any red flags. I‘ll just tell you from an operational perspective though, you know, I’m more old school than I might come across because we have most of our leadership sitting here in Seattle, for example, and we can go into the office. We do have an office here that’s, you know, half a fulfillment, Savannah for swag, but the other half is whiteboards because there is a lot of value to coming and collaborating in person.
[00:09:44] But there’s no reason to go into an office if you don’t have a reason to be there, just to sit in the aura of other people, it doesn’t make sense to us. So, we try and throw team events like happy hours and, you know, quarterly retreats, things like that to get people together. And, you know, people will come together and fly to an office, like, you know, flew to Austin recently to have a design jam with the team down there.
[00:10:03] Yeah. Especially now that things are opening up. But I don’t think there’s any world in which we all need to go to the same place and be restricted to a talent pool in one geographic region.
[00:10:12] Jon Penland: Well, I think you’re absolutely right that there is value in that occasional face-to-face time, particularly around strategic thinking.
[00:10:23] So, when you’re doing sort of that long-term strategy thinking or if you’re working on a particularly challenging or demanding problem. Right? So, they were like, there are situations where it’s really good if you can get everybody in a room to look at each other and have a conversation, but then on a day-to-day basis, you don’t need that.Right?
[00:10:42] Like on a day-to-day basis, everybody working from home works great. And Kinsta is a remote-first company as well. And one of the challenges we’ve had over the last couple of years is that it’s been very hard to find any of that face-to-face time and we’re just sort of coming, rounding the corner, being able to do more of that face-to-face stuff.
[00:11:04] And it’s just invaluable. So, even as like a, even as a remote-first company that believes strongly in allowing folks to work from where they are located, from where they work best that opens up talent pools and all, all of those things are true. It’s also really important occasionally to get folks in the same room. It sounds like that’s been your experience there at Archera, as well.
[00:11:28] Aran Khanna: I, I think so. You know, largely, especially at the early stages, you know, for things like brainstorming and long-term planning, things like that, it’s very much invaluable to get people in the same room to sync up and also to, you know, just build team trust.
[00:11:40] ‘Cause I think it erodes a little bit over zoom when it’s so sparse, right. And I think the other thing that’s been a pleasant surprise for me just in the last few months is the conferences are back and seeing, you know, particularly my sales team there has been fantastic. You know, we are kind of distributed on that side of the house, and seeing everyone in person I think helps, you know, build a lot of that camaraderie.
[00:12:00] Jon Penland: Yeah. It’s funny that you mention erosion of trust over Zoom, the way I think of it, there’s like this pyramid of building rapport with a colleague, right? Like, the lowest level is purely text-based communication. Right? Probably the lowest levels, like email, right? And then, you get to slack or something where you might have some real-time conversation, but it’s still all just text, typing back and forth.
[00:12:24]And then, like the next level in that pyramid is something like a Zoom call or a Google meet. But even at that level, you still haven’t reached what you achieve where you’re in person. And what I’ve found to be true, for myself, is that, particularly, folks who I might not naturally gel with the best that meeting them in person completely changes the dynamic of that relationship for the better. Right?
[00:12:48] Because a lot of times there can be little things about somebody’s mannerisms or the way that they work or the way that they communicate in writing that is just not working for me. And, and it’s causing our collaboration to not be super effective, to say it nicely.
[00:13:04] And I’ve found time and, again, meeting them in person I’m like, “This is not the person that I thought I was working with.” Right? Like, “I like this person so much more now that I’ve met them in person than I did when we were trying to work purely remotely.” Which to me is just one more sort of, I don’t know, weight on the side of the scale of needing to have those in-person connections periodically.
[00:13:31] Aran Khanna: Yeah, no, that makes a lot of sense. Yeah. I think our experience is probably similar. We have a bigger organization, so it’s every pair of people essentially there’s this dynamic. So, it grows and squared with the company.
[00:13:42] Aran Khanna: I think the more, even you can get teams together without you involved. It alleviates the pressure across the board.
[00:13:47] Jon Penland: Yeah. I’m curious about Archera’s origin story. So, I read a little bit on Archera’s website about founding Archera, but what I didn’t necessarily understand or see was like, what are the steps that led you to say, “Okay, it’s time to create a product and see if we can launch this thing out into the world.” So, can you tell us a bit about Archera’s origin story?
[00:14:09] Aran Khanna: Yeah. So, you know, in short, I think the experience that I had working at AWS, where I was part of a smaller company beforehand, we were working on this open-source, deep learning framework called MXNet. And that team was essentially acquired into Amazon to, to build out the SageMaker product.
[00:14:25] You know, we didn’t call it SageMaker at the time, but we essentially built that first version. And one of the biggest objections from customers was on the cost. And, you know, actually talking about the insurance, this idea that customers we were saying, “Hey, look, you don’t use GPUs all the time, but you’re buying them and sticking them under your desk with these big, deep learning jobs.” For the same cost that you’re spending on that one GPU you can do five jobs, which is really all your scientists are gonna do in the cloud at massive scale, way faster.
[00:14:50] And the biggest objection is like, “Look, I can’t even handle the commodity resources I’m using today. Why would I let the R&D folks loose on these, you know, extremely expensive GPU instances?” And I can’t even reserve them because it’s just such a bursty usage, you know, it’ll take maybe a month or two to train a job back then of those GPUs, just running until the think converges for, you know, even something like ImageNet because these were like the P2 instances, not as good as the GPU and all the fancy stuff that, that we have now, but even then, like big models take a while to train.
[00:15:19] And so, you know, I thought about like, “Hey, if you had, you know, no one’s training these all at the same time. If you have that insurance, you could really provide a smoothing function for this.” And we couldn’t get it done with an Amazon, obviously, there’s a ton of competing priorities when I started to think about this and talk about it internally. And I saw that all the other cloud vendors were gonna have the same issue.
[00:15:37] So, you know, seeing that gap and understanding kind of what the customer pain was and what that opportunity was. I think, you know, I actually linked up with our CTO, who is my brother, uh, and his background is in quantitative trading. He was a D. E. Shaw, doing algorithms for that and then he did pricing at Uber.
[00:15:53] So, he was doing the quantitative analysis for the algorithms that, that, you know, price rides in Latin America, essentially. And, you know, he basically said, “Look, we could actually make this a financial derivatives market. You know, there’s a way to actually price this and really make this workout.” And that’s really kind of the nugget of what sparked us to start building the first prototype, which we started, kind of incorporated the company May of 2019, had some, you know, hacks around before that, but then really started to, to land some startup customers with that easy, you know, click of a button, you know, insured, basically commitment strategy.
[00:16:26] And then, as that, as that grew thought that we could raise a seed route for ’cause the market was big enough. And, you know, the customers that we did work with were seeing a ton of value, right? And we had good revenue from them and all of that. And then, obviously, fast forward, grow the team, attacked more large enterprise customers like Fortive, which is a big Fortune 500 that we’re working with now.
[00:16:44] Some of these, you know, a much more meaningful, uh, kind of IT organizations then, you know, saw that we could, “Hey, turn this around and potentially raise the A.” And we’re on that journey right now to see if we can scale up our sales funnel because it was just me and, you know, basically, no one else doing those first few sales, even those first few enterprise sales.
[00:17:02] So, it’s nice to have a team now and really see if we can scale this thing up. So, it’s still, you know, early days relatively. Yeah.
[00:17:09] Jon Penland: What are some of the challenges that you faced or, or that you’re facing today? You mentioned scaling sales. Is there anything else that you would highlight where you go, you know, as, as we’re trying to scale Archera, these are some of the problems we’re, we’re thinking about today?
[00:17:25] Aran Khanna: I mean, obviously, as CEO, right, the biggest thing is people. Like, “Are we, you know, do we have the right hiring roadmap? How are we thinking about attacking this? And is this aligned with what the customers and the partners that we wanna work with are going to be demanding of us three months from now?” Because you got a plan for all that ahead and make sure you have the right leadership.
[00:17:42] I think we’ve got amazing people at our share. We’ve got, I think the best people in the space, honestly, that we could hope for. And, you know, particularly on the leadership team, we have some very strong leaders, but it took us a while to hire those people. And, you know, we have other people we wanna hire and it’s gonna take a very long time, you know, equivalently because I think, especially at the early days of the business, bringing in the first few leaders, you know, and obviously we made mistakes here in the early days, like back in 2020 and 2021.
[00:18:05] But, you know, that’s very painful for the business to get it wrong. So, I try to spend a lot of my time thinking about getting it right on that side. And then, you know, obviously, now we have a great VP of engineering, who’s, who’s got an amazing product sense, but making sure things are right on the product side and the vision and the roadmap is, is sort of all there.
[00:18:21] But, you know, I think between those two things, that’s what I spend a lot of my time on. And, and that’s what I think are, are the challenges that I see. Everyone’s working with their own challenges within the business.
[00:18:29] Aran Khanna: You ask my CTO on the technical side, there’s another set of very, you know, thorough challenges around risk modeling and things like that, that he has to take into account, but, you know, it’s exciting.
[00:18:39] Jon Penland: Yeah. I think hiring is one of those challenges that if you haven’t gone through that process before, you probably don’t appreciate how difficult that can be. It can be very difficult to, there’s just so many challenges tied up with it. How do you identify tasks that are, you’re gonna hand off? How do you craft those into a meaningful position?
[00:19:04] How do you identify the right folks? And it’s very easy to get that wrong. It’s very easy to, you know, not hire perfect person. And, and when you’re a small team, every wrong hire is amplified, right, it’s magnified. If you’re a team of 500 and you have a wrong hire, it’s one out of 500 people, even if it’s in a leadership role, if you’re a team of 20, it’s one out of 20, right?
[00:19:28] That’s very different. Are there any sort of, I don’t know, ground rules or principles that you use as you’re going through this hiring process to try and say, “Okay, these are sort of the things that are non-negotiables for us, to make sure that we end up with the right kind of team.”
[00:19:47] Aran Khanna: Yeah. I mean, I think a lot of it, frankly, has come just in the last few months as we’ve started to scale up a little bit more, kind of the headcount on the team.
[00:19:54] And a lot of this is, is due to our amazing, you know, HR partner, who, who’s helped us on taking, essentially, the leadership principles that, you know, coming from AWS, that’s something that I really thought was well implemented in the organization and led to a lot of positive outcomes on the product side.
[00:20:10] You know, obviously, it scaled, there were other issues, but as a core thing to align, you know, a disparate team, I thought that was very powerful. So, we started with having a course set of leadership principles that, you know, we’ve been tweaking over time. These things are never set in stone, but our HR contact really helped us hone that into a really a set of operational hiring practices, including questions and things like that that I think really helped us bring a more formal process. And especially as we had, you know, not just me and my CTOs sitting on, as the co-founder sitting on every single loop, but other people running loops independently to grow their segments of the business.
[00:20:43] It helped us basically delegate much more effectively in a much more aligned way. So, that’s kind of something that we’ve started to do now. And I think is probably, you know, from our perspective the best way to do principled hiring in this distributed world that we’re in right now.
[00:20:58] Jon Penland: Yeah, I’m curious. And I know I’m putting you on the spot here a little bit, trying to recall things for memory, but do you recall, off the top of your head, what some of those leadership principles are that you’ve sort of ingrained into the hiring process to drive that process?
[00:21:13] Aran Khanna: Okay. So, the biggest one is customer first, right?
[00:21:15] Okay. We are a fiduciary for our customers, right? So, that every decision that we try and make is driven by, you know, customer requests. Most of our roadmap is customer requests. And then, especially because we’re automating and sharing risk with our customers. Everything needs to be made in the best financial interest of the end customer.
[00:21:31] Right? There’s no world in which we would, you know, screw over one customer at the behest of another, for example. We have to be a fiduciary for each and every customer that we service. So, that’s just one principle that across the organization, I think, you know, operates as our top principle.
[00:21:47] Another one is extreme ownership, especially in the early stages. I think the ability for someone to say, “Hey, I’m just gonna stick to my lane, even though I see a fire burning over there.” You know, momentum can be really hard and in a distributed world momentum can be especially hard to sort of avoid. And something that we want to, you know, due to operationalize that customer-first mindset, you know, if the site is broken for example or, you know, data is showing up in a way that’s maybe not ideal is have anyone take, be able to take ownership over that. There’s no, like, guarding your lane, if something’s wrong for the customer, like you own it.
[00:22:20] If you’re the one who sees it and you can go and act otherwise, you know, escalate it to the appropriate places, but there’s no like, “Hey, I, look, I’m not going to, to act on something if I see something wrong.” And we try and empower people to that end internally and operationalize that, kind of in a number of ways throughout the company.
[00:22:37] Jon Penland: Yeah. I love both of those principles. The idea of whatever you choose to do as a company has to take the needs of your customers as a priority, right? Like those, those are the first set of priorities that you have to make sure you’re taken care of. And then, that second idea being extreme ownership.
[00:22:56] That’s something we talk about a lot, internally, at Kinsta about. I think the way that we have described it, a lot of times, is, “It’s not your responsibility to solve every problem you encounter. It’s not your responsibility to know every answer. It is your responsibility to make sure that problems don’t get ignored.” Right?
[00:23:16] So, if you have a customer and they have a problem, you don’t necessarily have to know the answer. You do have to make sure that that problem or that question gets an answer. Right? So, and that’s sort of how we’ve tried to ask our team to take ownership for their roles and, and really outside of their roles, really any sort of significant issue they encounter without saddling them with the responsibility for feeling like they have to know everything.
[00:23:44] Right? So, trying to find that balance between responsibility and realistic ownership of your area of influence.
[00:23:52] Aran Khanna: Yeah, I think you put that very well. I like the way you framed it.
[00:23:55] Jon Penland: Backing up a little bit from Archera. I, I’d love to hear more about your background before Archera. So, when did you first get into technology?
[00:24:04] Is technology something you were interested in growing up or is that an interest that developed in school or at what point did your interest in technology sort of, yeah, blossom?
[00:24:14] Aran Khanna: So, it’s kind of funny. You know, I think I almost boomeranged in a way to technology. So, I grew up in Seattle.
[00:24:19] I was born and raised here and obviously kind of grew up around the nineties, and Microsoft was growing and Amazon was growing. Actually, my parents worked kind of running these different teams within Microsoft and Amazon. So, kind of early days, I knew a lot about the technology industry. I had, you know, all the computers lying around the house that I could play with, got really into video games and things like that.
[00:24:39] But, actually, in terms of my interest, when I was in high school, I wanted to do biotech research. I was super interested in lab work and like my first internship ever was working at, you know, a local biotech company here in Seattle building genetically modified cyanobacteria, essentially in algae, that could basically produce free fatty acids, aka biofuel, from photosynthesis.
[00:25:04] So, a way to essentially decarbonize and, or maybe be carbon neutral, right, and turn that into kind of usable hydrocarbons. And, you know, the cycle times on that bioresearch was weeks. You know, I have to grow the alga, put it in and, you know, sometimes the fungus would contaminate it and I’m set back a whole quarter.
[00:25:23] And then, I started to meet, you know, especially as I went into college, more of the computational bio guys and their experiments were getting done in, like, five seconds. Yeah. So, I’m like, “What the hell am I wasting my life sitting on a bench doing?” And I started to dip my toe into computer science, and took some classes in college.
[00:25:40] And then, you know, just because I could build interesting things, just fell in love with it. And, you know, started doing tons of side projects, getting really deep into that. I majored in that, in mathematics in college, but, but then got really into kind of what’s the application to the real world. I think that was always something that excited me about computer science.
[00:25:56] I could make an interesting application. I could build something interesting that would actually have impact on the world immediately, rather than years down the line when we set up the, the bioreactor, you know, in the Pacific ocean or something like that.
[00:26:08] Jon Penland: Right. Yeah. Why, why am I growing algae at months at a time when I could be running an experiment where the computer does all the work, right?
[00:26:14] Aran Khanna: Exactly.
[00:26:15] Jon Penland: Yeah. Yeah. It did seem like while you were still in school, you did veer away from biotech into, as you described, how technology comes to bear on the real world. And some of the things that you, were involved in or projects that you worked on, I believe while you were still in school, had a strong bent towards user privacy, which is, I think, one of the primary places where technology collides with the real world today.
[00:26:43] Right? And, and the things I’m thinking about there was a Facebook messenger flaw that you discovered and we’re able to build a proof of concept, so to speak, and then there was an issue with how a Venmo shares information. Can you unpack those two different situations for our listeners?
[00:27:03] What did you find and what did you do about it?
[00:27:07] Aran Khanna: Yeah, it’s interesting. I mean, I could probably clinically talk about this from a researcher’s perspective, but I think the more interesting piece is, you know, how I got into it and kind of the story behind it. Because I was, you know, in school, studying computer science, I actually had this guest lecturer, or rather guest, guest professor in one of the classes that I was in at Harvard, who was the former CTO of the FTC. And actually worked on a lot of privacy-related cases.
[00:27:30] Aran Khanna: You know, definitely credit Latonya Sweeney. Dr. Latonya Sweeney was actually setting me down this path of thinking about these problems, but, you know, I was very passionate about privacy. Obviously, I used all of these apps on like, you know, some of the older generations, my generation was bored into basically communicating us with Facebook messenger, at the time, as a default.
[00:27:48] Aran Khanna: And communicating through social media, like Instagram, kind of with your, your wider friend group.
[00:27:54] So, those things kind of really started to, to get me thinking about, you know, what are kind of the unintended consequences of these, you know? We saw a little bit of that in privacy and technology, but I started to look at my own usage, like, “What’s going on here?” In Parallel, you know, like any good computer science student, got an internship.
[00:28:10] It was junior year or, you know, the summer before. Right? And I got an internship at Facebook, working on what was actually going to be the thing that I probably if I ended up working there, I’d have to go and be fired for writing another blog post. It was on the newsfeed ranking team. Which, obviously, if you can think about all the next wave of controversy for privacy, it’s very much generated by that.
[00:28:30] That being said, you know, I started looking at Facebook as a product and, I found that one of the interesting things on the mobile app was every time you sent a message via messenger, it would by default attached to the location of that message. Now, that wasn’t a flaw in so much as it was a weird design decision, right, where this thing is on by default and I can see the use case, maybe you’re messaging someone in the park and you wanna go meet up, so it’s kind of helpful for that. But, you know, frankly, what I saw was just anecdotally, I could actually see all these messages going in group chats and almost track people’s locations.
[00:29:02] Aran Khanna: Something I could do, you know, in my head with a pen and, you know, pen and piece of paper, I wanted to create an app for.
[00:29:07] So, I created this little Chrome extension called Marauder’s Map and this was kind of the summer before I was supposed to, right before I was supposed to start a Facebook, that would help people see essentially all this data that’s coming into their messenger feed, and plotted on a map. And, obviously, if you know Harry Potter, the Marauder’s Map helped Harry track all the people around Hogwarts.
[00:29:24] And people say that Harvard looks like Hogwarts. So, I thought it’d be cute to say, “Hey, I was tracking all my friends around Harvard the same way, you know, Harry was tracking people around Hogwarts.” It actually worked quite well, you know, funnily enough. But, you know, I put that as a blog post, quick open-source project, released it on the Chrome store and the thing basically went viral.
[00:29:42] And, honestly, I thought I was doing a service to the end customer on Facebook by saying, “Hey, maybe this is this something you should look at and turn off. They should, you know, my views, they should probably build this into their, to their privacy checkup, to see, “Are you exposing too much data here? Do you want to turn this default off?”
[00:29:55] But Facebook didn’t see it that way. And basically, the day before I was supposed to join, long story short, you know, after I deactivated the map extension and everything they had asked, they basically said, “Hey, you didn’t act in the best interest of Facebook.” And they fired me when I had an apartment and everything in San Francisco I had to pay for and was banking on that internship.
[00:30:12] Luckily, I landed at a startup, which was basically where, you know, the team that got acquired into AWS, that I was working for. But the whole experience basically showed me. And when I started my company, the customer always comes first. And when you’re at Facebook, you know, they have, I think kind of a toxic culture internally from what I saw, but also the customer is not the user of the product.
[00:30:33] It’s very clear. They’re not thinking customer first, the customer is the advertiser. And so, that’s what I really liked about Amazon. Very transparent about what the customer is and putting them first. And, you know, when I started my company, that’s why customer first is really the most important leadership principle that we have.
[00:30:47] So, that was kind of that piece of it. And then, I repeated the pattern with Venmo, basically, and worked with Latonya to release papers on all of these projects. Because Venmo had some invasive defaults for share all your, all your payments when you can, you know, track people’s friendships as they developed, you could see who they’re going on dates with.
[00:31:03] It was kind of fun, as well. But, yeah, uh, you know, the net of it and I did a TED Talk on this was that you could essentially see that in the early, in the early days with these platforms becoming popular, the privacy sharing was always flipped on by default because that was a lever for growth for them.
[00:31:18] And it just created a kind of perverse incentive for the end user. And without any regulation, I think the regulation still really hasn’t come far enough. The thing that’s been more, more effective than the Congress has been Apple at regulating privacy, which is funny, but you know, the incentives were really misaligned and we need someone to really help bring that back into balance where the companies aren’t incented to, basically, throw their user data out in, in the hopes that it’ll generate some growth.
[00:31:45] Jon Penland: Yeah. So, in those early days, you had this strong, demonstrated interest in user privacy, making sure that apps are clear about what are you sharing, making sure that apps are clear about where their priorities are, who their customers are. And I’m curious how that has translated into what you’re doing today at Archera?
[00:32:06] And you’ve spoken to that a little bit, but can you just speak to this commitment to the idea that users deserve to know what’s being shared and that company should be clear about who their primary customer is? How does that influence how Archera operates today?
[00:32:24] Aran Khanna: Yeah. I mean, maybe zooming out because I think a little bit is on the principle side, but I think the common thread that goes between my, my privacy research and what I was doing there from a, you know, technology across society standpoint, it really sent around to this idea of, you know, these big, powerful digital platforms essentially were really tough for customers to navigate.
[00:32:43] And the complexity caused a lot of downsides. So, trying to level the playing field for that end user of these powerful complex platforms it’s something that I think, in many ways, we’re trying to do at Archera today. When we think about, for any given AWS resource there are 36 ways to reserve it. And if you wanna say like, “Hey, for $500 that I could spend upfront in this basket of things I’m committing to, where do I put those upfront dollars?”
[00:33:06] That’s literally an NP-hard problem. Like, that is crazy that they embed NP-hard problems in planning your purchase of these resources. And, you know, obviously, it’s an efficiency that they, they soak up on the, on the profit line. But, again, the goal is to really simplify and make the platform transparent to that end customers so they know what’s going on and they can make informed decisions. And I think that’s the common thread in kind of the tools I build generally and what I like to work on, be it in the form of a company or in the form of an open-source kind of app and proof of concept.
[00:33:37] Jon Penland: Yeah. What caused you to decide to make the switch from pursuing this sort of corporate track that you were on as you were working with other, with major cloud companies, Azure, AWS, and at some point you decided to adventure out on your own, what transpired to decide to cause you to decide to make that change?
[00:33:57] Aran Khanna: I mean, I think, frankly, it was the opportunity that I saw to make customers’ lives a lot easier by solving this problem that I ran into and just couldn’t see a line of sight to solving inside a big company. You know, there, there is a natural intransigence to a big, leading vendor of any space.
[00:34:17] And I think, you know, especially while I’m, you know, while I’m young and spry, I want to be able to, you know, channel that energy into something where, you know, the marginal thing that I do actually has marginal impact, even if it’s on a smaller scale.
[00:34:31] Jon Penland: So, I mean, what I’m hearing is the motivation is more in the direction of wanting to feel proud about the work that you’re doing, as opposed to necessarily trying to build, you know, a unicorn or, you know, hundred million dollar valuation. Although, I mean, you can have both quotes, right?
[00:34:51] Aran Khanna: I’m trying to solve a problem for customers.
[00:34:53] Jon Penland: Right.
[00:34:53] Aran Khanna: I’m excited about solving a problem for customers. And I think this is something where, if I put in a year of work, I can move the ball way more, you know, way further down the field for customers and probably in the long term, move it further down the field for more customers than I could when I was within Amazon or within Microsoft, just because of, you know, the layers that you have to go through to affect change there.
[00:35:13] And that’s kind of the bet I’m making. I can make more of an impact and improve more customers’ lives doing this. And I’ll obviously feel more proud about having more of an impact.
[00:35:20] Aran Khanna: But I think the motivating thing is that this is a way to, much more quickly in my estimation, move the ball, the ball down the field for customers.
[00:35:27] Jon Penland: Yeah. You’ve been in this startup space for a couple of years now. And as you think about folks who might be coming down the track behind you, is there any advice you would give to somebody who maybe is right outta school or who maybe isn’t, you know, with a cloud provider right now thinking about leaving and launching their own company, are there any lessons that you’ve learned that you’d share or any pitfalls that you would warn others to avoid?
[00:35:52] Aran Khanna: Yeah, I mean, tons, but, uh, that, that, that would be us sitting here for, for hours. I don’t know if we have too much more time. So, I think kind of broad strokes, what was helpful to me on the journey was, you know, particularly talking about earlier stage people who are maybe in college or coming out of college.
[00:36:09] Aran Khanna: Using those early years to get a really good breadth of what’s out there, maybe working at different companies.
[00:36:13] And, honestly, like, you don’t have to do a startup immediately when you come outta college. In fact, in many ways it’s better to go and find, do your exploration through your internships, find a kind of company that’s interesting, be it a cloud provider or whatever, right, in whatever domain. And operate there like an owner, you know, operate there, take on more than, than you are, you know, and you obviously have to land in the right team, there’s politics and other stuff at organizations.
[00:36:38] But if you can go to a relatively nimble organization, maybe a late-stage startup or something like that, and really learn in a space that you’re excited about building you’re gonna see way more, get way more customer connections and be really set up in, in an interesting way where you have almost founder market fit to go and start something.
[00:36:55] So, that’s something that I would, you know, just generally is a piece of advice to someone in college, give, if they’re interested in going down the entrepreneurial track in the long term. Obviously, there’s many exceptions to that, but that was my general, you know, the enterprise SaaS space, particularly in the cloud space, my experience.
[00:37:10] If someone’s looking to leave and start something, I think the, there’s many ways to de-risk it and whatever, but the easiest way is just to start. You know, if you take the leap, it’s hard to, you know, if you go through a one-way door, it’s hard to not commit a hundred percent of yourself to that endeavor.
[00:37:28] So, you know, you can work on in, on the nights and weekends, you can hone the idea, but, at some point, you have to, you have to make the leap and, and actually go build the thing and validate it with customers. And that, if you’re doing it right, I think should be a full-time job.
[00:37:40] Jon Penland: Yeah. Yeah. And I, I really love one phrase that you mentioned in there, which is operate like an owner.
[00:37:47] And I think is, kind of gets back to that idea of ownership, of this ownership mentality that you’ve mentioned and that we make such a big deal of at Kinsta is that. And there’s a way you have to caveat this a bit because when you operate as an owner you have to also understand the limits of your decision-making discretion.
[00:38:05] Right? So, so, when we talk about operating as an owner, it’s not making decisions necessarily that you’re not allowed to make, it’s taking responsibility for the full breath of what you encounter. Right? Taking, taking responsibility, having a desire to drive success for the organization, not just focus on specific responsibilities.
[00:38:25] I feel like that’s advice that anybody, even if they’re not, you know, interested in starting a company, right, like that’s just great career advice in general, is this operate as an owner.
[00:38:34] Aran Khanna: Totally. And, honestly, like, I think if you’re at a high-performing company that recognizes and, and gives more to people who are leaning in, you’re gonna see way more of the business.
[00:38:43] You’re gonna have more scope because you’re going to say, “Hey, I wanna be in the meeting with the customer.”
[00:38:47] Jon Penland: Yeah.
[00:38:47] Aran Khanna: “I wanna be kind of on the phone as we’re onboarding this person. I wanna actually sit in on, you know, a customer success call. I wanna learn.” And, you know, if you’re in, in a culture that and I think, you know, you should strive to be in companies that have the culture that support that.
[00:39:02] Aran Khanna: And give you that scope. Because I think it’s great mutually. You’ll benefit the business more, you’ll know more and you’ll be able to make better decisions and, you know, that benefits the business, but you learn a lot more if you lean in and operate like an owner. So, I think it’s mutually beneficial.
[00:39:17] Jon Penland: Absolutely. So, as we wrap our conversation to a close, I have two close-out questions that I always ask all of our guests. So, the first is what is a resource that you would recommend to listeners of this podcast? This could be a book. It could be a blog that they should follow, a newsletter, really anything is a good answer. What’s something that you value in your own life that you’d recommend to our listeners?
[00:39:42] Aran Khanna: Well, I mean, topically, kind of staying consistent here. The book Extreme Ownership is one that I would definitely recommend, it was written by a former Navy seal. I’m sure that you’ve read it or heard about it, given that I think a lot of these principles and, and these ways of operating, I think come from that book and obviously in the most extreme conditions, what a culture like that can enable people to do. And it’s, it’s quite fantastic. So, definitely recommend that. And then, maybe another plug, I love the Acquired podcast, if you’re interested in entrepreneurship.
[00:40:11] Aran Khanna: There’s like three hours, hour and a half deep dives into businesses.
[00:40:15] Actually, one of our seed investors, Pioneer Square Labs, one of the partners there is one of the hosts of the show. And I think they just do a fantastic job. If you’re interested in business and entrepreneurship, it’s a very interesting long-form podcast that’s worth, worth to listen.
[00:40:27] Jon Penland: Awesome. We’ll make sure we get links to both of those in the show notes. And, uh, the last question for you. If somebody wants to connect with you or to learn more about Archera, where would, where should we send them?
[00:40:39] Aran Khanna: Yeah, you can find us at Archera, A R C H E R A.ai, uh, and you can find me on Twitter at A R A N K H A N N A.
[00:40:49] Uh, and I’m usually tweeting about Archera or you know, obviously feel free to DM me if, anything in this episode is interesting to you and you wanna have a further chat.
[00:40:59] Jon Penland: Yeah. Well, Aran, thank you so much for hanging out with me today on Reverse Engineered.
[00:41:04] Aran Khanna: It was great to be here.
[00:41:05] Thanks so much for having me, Jon. This was a lot of fun.
[00:41:07] Jon Penland: Yeah. And thank you to our listeners. That’s all for today’s podcast. You can access the episode show notes at kinsta.com/podcast. That’s kinsta.com/podcast. If you enjoyed this episode, don’t forget to subscribe to Reverse Engineered and leave us a review on Apple Podcasts or the platform you’re listening on right now.
[00:41:27] See you next time. [00:41:29] Outro: Thanks for joining us on this episode of Reverse Engineered, the podcast on all things business and tech. Reverse Engineered is brought to you by Kinsta. Kinsta’s premium WordPress hosting can speed up your website by up to 200% and you’ll get 24/7 support from our expert WordPress engineers. Let us show you the Kinsta difference at kinsta.com.