Paul Lappas Discusses Intermix and Mission-Critical Data

Paul Lappas of Intermix.io - Podcast Transcript

Scott:	Welcome to Founders and Friends podcast. This is Scott Orn at Kruze Consulting. My very special guest today is Paul Lappas from Intermix. Welcome Paul.
Paul:	Hey Scott, thanks for having me. Excited to be here.
Scott:	Paul is a CEO we’ve worked with … You guys have been a client for two or three years now, right?
Paul:	Yeah, I think over two years.
Scott:	Yeah. His partner is named Lars, who’s a great guy as well. It’s been awesome working with you guys. Finally, we were like, “We’ve got to have you on the podcast.” You made it happen. Thank you for coming over on a sunny Friday.
Paul:	Well, you know, we’re only a few blocks away, here. It was literally closer than going to get lunch today.
Scott:	Nice. We have LaCroix and coffee for you, so it’s a double doozy.
Paul:	Two of my favorite things.
Scott:	Well, hey, so Paul, tell us about Intermix. Retrace your career, and how did you come up with the idea for Intermix?
Paul:	Well, yeah. I’ve been in the valley for almost 20 years now. That makes me sound really old.
Scott:	You look way too young to be saying that.
Paul:	I know, well thank you for that.
Scott:	Yes.
Paul:	Earlier in my career, I had co-founded one of the first cloud computing companies. I started with infrastructure technology. Then at some point I realized, “Wow, I really hate wrangling servers and being on call.”
Scott:	Like the midnight wake up because something’s down?
Paul:	Yeah. I spent the majority of the first ten years in my career answering support requests in logging into server terminals while I was drunk. On Friday night, you still have to go out, right?
Scott:	Yeah. You’re not a doctor. You’re not operating on people. This is a computer your working on.
Paul:	It’s people’s websites that are down. That’s okay. Then in 2012, ‘13, landed at a company that was doing mobile, mobile crash reporting. Their name was really interesting. The name was Crittercism, which I-
Scott:	Oh, I’ve heard of them. Yeah, yeah.
Paul:	Okay. We found the critters in your mobile apps. When I had joined, they already had signed up a ton of the large mobile brands, like ESPN, NBC. They had a ton of users. Yahoo. They were trying to get revenue. That’s where I met Lars. He joined as the head of BD. I was the head of engineering. We started selling the service. It was a crash reporting solution for companies so that they could find bugs in their apps. We were sitting on a whole bunch of data. For example, we had a library, a little piece of software installed in over one billion devices all around the world. We were receiving-
Scott:	That’s a billion with a B.
Paul:	Yeah.
Scott:	Yeah.
Paul:	Across Android. At the time, Microsoft still had a mobile OS. Obviously, Apple. Our platform was receiving like 60 thousand events per second from all these apps. Like metrics and telemetry. We were selling this service and trying to grow that business. Then one day Lars approaches me like, “Hey, I’ve got some folks who would be interested in buying this data.” I was like, “Okay. I’m not really sure who this guy Lars is, but he seems like a nice guy.” I’m like, “Who?” He named this Fortune 100 company. He was like, “Do you think you could put it somewhere whereby we could give it to them and they could access it in their environment?” I was like, Okay, sure.
Scott:	When they want to see usage, and geography and …
Paul:	Exactly. For example, we had data that could inform you of something like the following. We could tell you how many activations occurred in January on ATT’s network on this specific Samsung device in LA. We had all that information across all these mobile apps and every single device that was out in the world. I was like, “Sure.” What I ended up doing is I hired a data scientist. It would be this person’s job to go and get this data ready. Looked for a data scientist, and these people are kind of expensive. Not really a cheap role to hire for.
Scott:	I’ve heard that, yeah.
Paul:	First day on the job, the guy was like, “Great, where’s the data?” I was like, “It’s in the servers, right here.” He’s like, “Well, I can’t really use it. I need it to be in one place so that I can see all of it, have access to all of it, overall time in a way that’s clean and correct and complete, and be able to use my tool of choice.” Data scientists have special …
Scott:	They’re like a painter that needs the right pigments, and canvas.
Paul:	The delicate genius, right, needed to use his own tools. He needed that before he could do any work. Then I was faced with the choice of like, “Okay, what do I do? Do I basically let this guy go and then go fix that problem?” I said, “No, because it was hard to find him.” Basically, he was sitting idle for like three months. I was paying him a salary but he wasn’t really doing that much. I had to go off and build another team that would actually get the data to a place where it was useful, so we did that.
Scott:	This is becoming a more and more and more expensive project, here.
Paul:	Yeah, yeah. It was not what I signed up for. We ended up hacking something together. We actually made that deal. We actually charged millions of dollars for that data and someone paid for it, a few companies paid for it. In doing that, I started chatting with a bunch of other peers in the industry and figuring out how they solve problems like this. It was like the Wild Wild West. Everybody had the same story. There were a lot of different ways that the problem was being solved and everybody was struggling. When I ended up leaving that company, Lars left a little bit after I did, we said man, we didn’t really want to go back to work for someone else. Why don’t we go and see if we can solve some problems in this area? So we did. That’s how we started Intermix.
Scott:	Give the Intermix pitch. What do you guys do?
Paul:	Okay. We actually honed this a few months ago after our latest round of funding.
Scott:	Did you hone it before or after you raised the money?
Paul:	After. Afterwards. Intermix is a single dashboard that helps data engineers keep an eye on their mission-critical data flows. When companies want to be data-driven … Look, it’s no surprise, you ask anyone, any company, any owner, or any leader of a company, “Is your company data-driven?” The answer is, “Yes.” Any earnings call, data, artificial intelligence, is constantly being referenced. The fact is that data is now more important than software.
Scott:	Yeah, because it’s kind of like what you get out of the software. I’m not a developer, but that’s how I think about it, like the finished product, or the signals.
Paul:	Yeah, yeah. If you’re making complex decisions based on how your customers have interacted with your product or basically anything that you want to help add value to your products or to your business, traditionally you’ll hire a software engineer to code a bunch of if-then statements into your application for you. Say, “If this happens, then that happens.” That’s not really the best way to make decisions at a large, large scale. It’s kind of like you have a nine-month-old daughter, right?
Scott:	Yeah.
Paul:	She’s going to learn to walk at some point. I don’t know if she’s walking yet.
Scott:	She’s very close. It’s a little scary.
Paul:	She’s probably knocked her head a few times, or she’s fallen down.
Scott:	Yeah, it’s a data-driven decision not to knock your head.
Paul:	She’s realized, “I should not do that. I should stop knocking my head.” She sort of learned that was what she shouldn’t be doing, because it really hurt when she hit her head. In some level in her brain, now, it’s been programmed in. Some software has been written, essentially, that basically says, “If I do this, it’s going to hurt.” That’s what machine learning helps you do. It helps you use all of those little experiences about how your users interact with your product. All these little data points all add up, so that you can ultimately make a better decision about the future. You can never program that. The if-then statement list would just … The recipe would just be too long.
Scott:	It’d be too much, yeah, yeah, yeah, yeah, yeah.
Paul:	Yeah. You can think about it in a way that data is writing software, now. Companies-
Scott:	Data is writing software through machine … Machine learning writes its own …
Paul:	Rules.
Scott:	… Based on data, yeah, yeah, yeah.
Paul:	Yeah.
Scott:	That’s actually a really cool way of saying that. I’ve never heard anyone say it that way before.
Paul:	Actually, that was the first time [crosstalk]
Scott:	I need to process that for a second. Once you get the good, clean data, the machine learning can really work its magic and write software that takes advantage of that. That makes perfect sense.
Paul:	If the examples are true, if the format of all of them is the same, and these algorithms can actually run on it, every single time an example comes in, your brain, your model, will be trained one more tiny incremental step. It’ll get that much better in predicting the future. It all depends on data. That becomes the core IP now for a company. The data that a company has is proprietary. Any model that it builds on that data it now owns.
Scott:	Yeah. By the way, we, at Kruze Consulting, are doing this too. We have an amazing CTO, [Brent Ballard] , who’s writing … We’re pulling a ton of data now, and helping us make decisions on when to talk to our clients about certain services, things like that. Even like an accounting firm that … You guys probably don’t even know we do this yet, but we’re pretty progressive. We’re doing it. Everyone kind of needs to think this way. I just interrupted you.
Paul:	If you don’t, if you don’t, then a competitor’s going to come in and do it.
Scott:	Yeah, do it better.
Paul:	They’re going to ultimately build a better product for their customers. Then you’ll be out of business.
Scott:	Yeah.
Paul:	That exact story is playing out all over the world, right now. Every single company, via a tech company, a new tech company or an old-school company, has sort of realized that that’s true. The reason why that’s happening now is because of the cloud. The cloud makes it accessible. Anybody can store all of the data forever.
Scott:	I was going to say, storage, right.
Paul:	It’s cheap. It’s cheap, right?
Scott:	Yeah, and you can process it. Not just store it, but process it, right?
Paul:	Yeah. You can process cheaply on a just-in-time basis. You don’t need to put up a bunch of money to process it, upfront. You can just do it as you go. Number two, and a lot of people don’t know this, but machine learning algorithms, there’s only like seven types. There’s not that many types of algorithms.
Scott:	Oh geez, yeah.
Paul:	They have to be crafted and tuned, and all that. The core math that drives them is about 70 years old.
Scott:	70?
Paul:	Yeah. Now, it’s been codified into these opensource libraries. Have you heard of something like TensorFlow?
Scott:	I have heard of TensorFlow, yeah.
Paul:	That is basically the codification of some of these really old algorithms. Now it’s open source and anybody can use it. You can use it. I can use it. But it’s useless without data.
Scott:	Yeah. My reference for a lot of this stuff is a documentary I saw on the one that Google acquired, the British company, that did the chess … Was it chess? I think it was chess?
Paul:	Recently?
Scott:	Yeah.
Paul:	Go? Alpha go [inaudible] ?
Scott:	Yeah, yeah, yeah, yeah. Whoever did that.
Paul:	Google.
Scott:	I was watching those guys. Yeah. I didn’t know these are really old algorithms. So what happened? Has the processing speed just gotten so much faster and the data sets are bigger, so that allowed you to basically make these old algorithms work? Is that kind of what …
Paul:	Well, what happened is these data scientists at the large tech companies like Google and Facebook went off and codified these algorithms in modern programming languages, and made them available to executed on cloud computing clusters. Then they just open sourced them.
Scott:	That was really nice of them.
Paul:	Yeah. In fact, the history of technology is just littered with large companies open-sourcing software that gets leveraged. Ultimately, that’s not their business model.
Scott:	Yeah.
Paul:	As an aside, but related, the algorithm that Google uses for page rank was actually invented in the late 1800s.
Scott:	My mind’s blown! What?
Paul:	Larry and Sergey, they actually were reading a research paper on the page rank algorithm and said, “Oh, we could use this to classify and rank the importance of web links.
Scott:	Yeah, real fast, this is a good digression. What was it being used for in the 1800s? Like library catalogs?
Paul:	I think it was just some guy … No, I don’t think it was used at all. I think it was just some dudes-
Scott:	Like theoretical.
Paul:	It was some dude’s PhD dissertation, or something.
Scott:	Wow. My friend’s dad was their computer science professor in Maryland, or one of the guys in Maryland. He called his son up, who’s my buddy, we work together, like, “You might want to join these … “ He would have been the first business employee at Google. My buddy was like, “Nah, whatever. I’m doing the best in banking.”
Paul:	It’s a stupid name.
Scott:	Yeah, yeah, yeah. Anyways, so everyone’s codifying the old algorithms, but they work better because it’s in the cloud now, and you can [crosstalk] -
Paul:	It’s no secret sauce, anymore. Any company has access to the same algorithms. What’s happening is companies are building these data teams. They’re saying, “Yep, I want to be data-driven. Yep, I’ve got to build AI,” for the same reason that you mentioned. For competitive advantage; bottom-line, top-line optimization; build new products; win in the market. So they’re forming data teams. Oftentimes, these data teams report right to the C-Suite. They have new budgets now. It’s not part of IT, and new roles. Data scientists, data engineer, it’s a new role [crosstalk] .
Scott:	Sounds like someone should be selling them a really awesome tool to manipulate that data.
Paul:	Well, if only anybody could solve their problems, they would make a lot of money.
Scott:	The new budget thing catches my eye, actually. Building a startup, selling into that new budget is really nice. That’s a good place to be. People have money, they’ll try new things. They have to show results to their bosses to justify that budget, so they’re wanting to experiment.
Paul:	You’re exactly right. As we have grown our business, we started our business focusing on … If you go to our website, you’ll see some of the companies that we work with, but middle-tier, late stage, VC funded tech companies and smaller companies were our customers.
Scott:	In my world, though, I don’t know if you can disclose your client list, but I see the names of your … When you issue POs to them, or invoices to them, you have the who’s who of late-stage or newly public companies that you’re working with. It’s very obvious. I look at your … And I’m like, “Oh, these guys are super aggressive. They’re growing so fast and they probably have problems. Intermix is solving these problems.” It’s a pretty badass customer list.
Paul:	Thank you. We love working with them. Yeah, they gave us permission to put their logo on our website, but I don’t think I can release it here. Let’s just say there’s one really large real estate company that is selling real estate, that is carving up real estate and selling it for co-working spaces that’s our client. They’re really a data company. People think of them as a real estate company, but they have sensors all over their buildings. All that data, how people move around the office-
Scott:	That’s fascinating.
Paul:	… How many times you go to the bathroom, what hallways are the most populated, all that information becomes this body of analytics that they leverage to figure out how to better design the buildings and all these things like that.
Scott:	Yeah, they just use the space more efficiently, right? We actually use them in San Jose. I think we’re going to end up using them again in a couple of other places. The value prop is actually really good. Probably because they’re just using their resources more efficiently. It’s that simple.
Paul:	I think so. You have your own office. [crosstalk]
Scott:	Yeah, you’re sitting in our own office, where we have-
Paul:	It’s hard to run an office.
Scott:	It’s hard. It’s expensive. We have 30 people in this office. Candidly, I wish I would have used that company. I wish we hadn’t had this big office. It’s kind of like the, probably what people used to say, it’s a total digression from what your real business is.
Paul:	That’s right.
Scott:	I wish I didn’t have to deal with it. Me and [Tatiana] are the ones that have to deal with it. No one else.
Paul:	Right.
Scott:	Anyways, yeah. That’s a good example of a company in the real world who’s probably in a sleepy industry that no one thinks about, using data and your product to be more efficient.
Paul:	We’re also seeing, recently we’ve started getting some enterprise clients, too. In the enterprise, the story’s a little bit different. Really big, old companies, the way that they innovate is they build new teams and they firewall them off of the old teams. Old folks, you can’t really teach an old dog new tricks, usually. That’s what we see a lot. We have a client out of a huge supply chain, 18 billion dollar company in Hong Kong that owns 14 thousand clothing manufacturing facilities. They use their data infrastructure to predict fashion trends. They’ve written all these crawlers. Apparently, if you follow about 100 different influencers on Instagram, you know all you need to know about fashion, so they crawl that. They crawl Google searches. They do image recognition on what people are wearing in these Instagram photos. They classify them using, “This sleeve is new,” or like, “this person’s wearing a white handkerchief in their shirt.”
Scott:	Wow. You can tell.
Paul:	They put all that into a machine learning model to try to predict what’s going to be hot next. Their ultimate goal is to optimize their supply chain so that they can pre-purchase the right textiles for all these 14 thousand manufacturers.
Scott:	That’s always been the hard part in fashion, is getting what people want to market fast enough.
Paul:	Yeah.
Scott:	That’s interesting. I have a former Kruze client in a company that I invested in, I’ve been friends with for a long time, Nadine West, does something like that. They use technology to predict fashion trends, and they’re on fire. I don’t think they’re doing it in the same exact way, but it actually makes tons of sense. That’s crazy. Fourteen thousand facilities? That’s a lot of data.
Paul:	That’s a lot of data. It’s a really old company. They’re using us, too. It’s super … The thing that I love the most is just hearing all the stories about what people are doing with data. I think you talked to five different companies. The core team is doing the same thing. They’re just crunching data, but what they’re doing with it is different.
Scott:	How does Intermix make it easier for them? What’s the pain point you guys solve?
Paul:	Yeah, so companies hire data scientists. Those data scientists are really good at running algorithms on data. But like you said, you’ve got to have the data for them to do it. It turns out that getting data to a place where it’s useful is really, really hard. That job falls to a role that in new company, that’s called a data engineer. In older enterprises, their still called DBA’s. That data engineer, they’re really sort of the janitor of the data world. What they do is, it’s really an un-sexy job, but it’s incredibly difficult. The data that we’re talking about is basically data that is either generated by your application, like how your users interact with your mobile app or your website. It’s also information that is contained in your customer relationship system. It’s industry information. Like in this example, “What is the cost of this textile today?” This data, it resides outside of your cloud. Some in, some outside. It’s all in different formats and it’s all over the world. Somebody has to bring it all in, move it to the exact same format-
Scott:	That’s tough.
Paul:	… And aggregate it. Crunch it, clean it, and then move it to a data warehouse or to a set of databases that are compatible with the programs and the tools that the data scientists are using. And, be able to document it in a way where it makes sense, what it is.
Scott:	Yeah. It’s like the Tower of Babel problem. You guys are taking all of this disparate data, making it the same format, then putting it together, then making it accessible.
Paul:	Exactly right. When you add to that that the volumes are constantly growing, fresh data is much more interesting than old data, so they expect it to be updated once a day, four times a day. It’s exactly the same as building a car. We call it a data assembly line, where there’s just dozens or sometimes hundreds of little steps that the data goes through before it gets to its final destination. If it gets stuck anywhere along that route, then the data’s not fresh, the data’s not complete, and the data’s not available to the data scientists. That’s what we help to eliminate [crosstalk] -
Scott:	Yeah, and those are expensive people. Their opportunity cost is pretty high. So you want to have it as fast as possible to them, right?
Paul:	Yeah, efficiency’s a big problem. When things do get stuck, they really don’t know what the reason is, or where it got stuck, or how to fix it. We help them with that, too.
Scott:	I think you guys have a big affiliation with Amazon, right? What are some of the platforms you guys work with or use to make this possible?
Paul:	So Amazon Redshift is a big one. Amazon Redshift is the data warehouse offering for Amazon web services. I think now it’s generating over a billion dollars in revenue to Amazon. It was the fastest growing service in Amazon’s cloud. We’re branching out next year to Snowflake and other data warehouses.
Scott:	That’s a good one. I see a lot of storages in Snowflake.
Paul:	Yeah, we’re going to continue to do that. That’s what our plan is for next year is to continue to expand out. Look, the whole market’s growing, but we have to go where our customers are taking us.
Scott:	So what does Snowflake do that’s different than Amazon? Are you just …
Paul:	Yeah, good question. Snowflake does something pretty interesting. They do two things that are interesting. Number one is they separate storage and compute. With Amazon Redshift, when you’re a customer and you have more data that you want to add to it, you need to scale it up. You pay more as you scale it up. When you do that, you basically have to purchase these blocks of computers that combine processing and storage together. Sometimes, you don’t really need more processing power. Sometimes you just need more compute power. But they force you to buy it all at the same time. The benefits of that is that your monthly bill is pretty static, because you’re just buying these blocks. Snowflake took a different route where they said, “We’re going to take a different route where we’re not going to make you buy these blocks. We’re going to charge you for every single query that you run.”
Scott:	So the more you use it, the more you’ll pay?
Paul:	Yeah, yeah. They have other issue there. I’ve heard stories where people run a really big query on a Friday. They come back and they have like a thousand dollar bill, or something, because that query ran for 12 hours.
Scott:	Where does Intermix fit in with Redshift? How do you guys compliment Redshift?
Paul:	Well, you know, Redshift is the core of the data team. That’s where all the jewels are. That’s where all the applications connect to. There’s three kinds of applications. There’s applications that get data out of Amazon Redshift to do reporting or to run machine learning algorithms, or for applications. Then you’ve got to get data into these things, too. When you launch Snowflake or Redshift, it’s empty. You’ve got to get data into it, somehow, and that’s another-
Scott:	Oh, I never thought about that.
Paul:	Yeah, it turns out to be another type of application, that’s running jobs on the system. Then you have a third type of application that is crunching data. Any single Amazon Redshift warehouse that we work with has over 10 different applications that are connected to it, and thousands of humans that are in one way or another depending on data that resides inside of it. What our tool does is it gives you a way to automatically discover all of that complexity. We will tell you, “Look, these are the 10 apps that are being used. These are all the people that are running jobs. By the way, here is how they are ranked. Here is the most expensive ones-
Scott:	Do you let them prioritize or give clearance?
Paul:	No, because we set off to the side. We’re not in line of the data, on purpose. We don’t … It’s a whole different-
Scott:	Yeah, yeah, yeah, yeah. That would be like a failure point or something.
Paul:	Yeah, exactly. We don’t want to do that. We will tell you a few things. We will say, “Here are the applications that are having issues with performance. Here are some users or some applications that are really expensive and resource wasting hogs. By the way, if there’s a problem, A, there’s a problem, and B, this is the cause of that problem.”
Scott:	That’s awesome.
Paul:	We give you all that visibility that before you were really, really flying blind. With Snowflake, we’ll also help you figure out cost attributions. Companies are spending millions of dollars a year on these databases. You kind of want to know, “What percentage of that bill can I contribute to different departments in my company?”
Scott:	Sure, and was it an ROI positive project, or something like that, or someone’s pet project.
Paul:	Exactly. For the record, it turns out that sales and marketing usually are the biggest users.
Scott:	Really?
Paul:	Yeah, yeah.
Scott:	Interesting.
Paul:	You know, because they want to see metrics up to date and all that.
Scott:	People, they must love you, because of the flying blind thing. Do you hear a lot of stories about people having a massive Amazon bill, or …
Paul:	Yeah.
Scott:	I didn’t know about Snowflake having huge bills like that, but definitely see it with Amazon.
Paul:	Yeah.
Scott:	Do they see you as a cost-containment thing? Or are you a performance enhancer? How do you position yourself?
Paul:	The smaller ones definitely see us as a tax on Amazon Redshift. If you’re a cost-conscious company, they kind of see us that way.
Scott:	I would think they would, the cost-conscious ones would love you, though. You’re telling them if they’re wasting money or not.
Paul:	True. Oftentimes, they will spend less on Amazon Redshift after they use us, so we’ll pay for ourselves.
Scott:	Yeah, yeah. Yeah, exactly.
Paul:	The bigger enterprises don’t see us that way at all. They see us as a way to make a team more efficient. If we can make their team 10% more efficient, then-
Scott:	Yeah, if they’re spending $2 million dollars on data scientists, that $200 thousand in savings.
Paul:	Exactly. More and more as data becomes more mission critical, then there’s also a calculation on what does it cost the business if that data’s down, or if the data’s not available?
Scott:	Yeah. Yeah, yeah, yeah, yeah, yeah. That’s fascinating. This is all for the pain point you had at Crittercism where you’re trying to aggregate all this stuff and clean it up. Now you just built it, that’s awesome.
Paul:	Yeah, yeah. Yeah. It’s been a long journey, Scott. We started giving software away for free, actually, when we first started.
Scott:	I didn’t know that.
Paul:	Yeah.
Scott:	You guys had revenue really quickly when you came to us, which is always a good sign, actually. For people who are curious of like … We have 170 monthly clients, so we see the ones that do well and the ones that don’t. Usually early revenue is one of the best indicators you could ever have. Even if it’s $10 thousand or $50 thousand, it shows that there’s a pain point and someone’s willing to pay for it. It’s a really big deal.
Paul:	We made a choice to do that early on. I think there’s two schools of thought on that, especially with hard tech companies like us, is let’s lock ourselves in a room for two years, take a bunch of VC money, then we’ll have a big launch and hope that somebody uses it.
Scott:	I’m not a big fan of that model. That’s a bad model as an investor.
Paul:	No, no. I came up through product, as well. I don’t like to write one line of code unless I know that on the other end of that, I’m solving a problem.
Scott:	Our CTO is like you. It’s good. He always forces me to tell him how much money we’re going to save or make in revenue for anything he’s doing.
Paul:	He’s a keeper.
Scott:	Yeah. He’s good.
Paul:	Yeah.
Scott:	Yeah. I would just be like, “Yeah, build it. That sounds awesome, let’s do it.”
Paul:	Well the problem with software, too, is once you build it, if there’s a feature that you build and release and it’s not being used, you end up still having to support it. It’s still this thing that’s there. It’s just demoralizing for the team, as well.
Scott:	This has been fascinating. Where are you guys going? What’s your next big … It sounds like Snowflake is the big one.
Paul:	Yeah, we’re going to support Snowflake this year.
Scott:	When is that going to come out, or is it too early to tell?
Paul:	That’s going to come out at the end of Q1 next year.
Scott:	Okay, cool.
Paul:	Yeah, yeah. We have a number of early customers that have committed to working with us on it as Alpha users. That’s-
Scott:	I’m sure Snowflake’s excited about that too, right?
Paul:	They are. We actually have talked to their leadership. They totally get what we do. They said, “Look, we’re not going to build any of this stuff. We’ll support you in the getting the word out to our customers once it’s ready.” So we’re excited about that. That’s next year. Our goal is to continue to grow and keep our customers happy. I think if we focus on that, we’ll be okay. Day-to-day.
Scott:	Keep doing what you’re doing. Yeah, yeah, yeah.
Paul:	Living the dream.
Scott:	Yeah, exactly. Well, maybe, tell everyone, reiterate Intermix and where can they find you?
Paul:	Sure thing. Thank you, yeah. We are Intermix, www.intermix.io. We are a single dashboard for data engineers to keep an eye on their mission-critical data flows. If you’re using Amazon Redshift or Snowflake, you have a large team, and you’re using data for business-critical purposes, reach out to us. We have a free trial. You can sign up on our website without talking to anyone to get started really, really easily.
Scott:	Yeah. I can vouch that you guys are doing … It’s exciting to see your revenue ramp and the kind of customers you have. The dogs are eating your dog food. It’s pretty exciting.
Paul:	We love working with you too, Scott. Thank you.
Scott:	Thanks man. All right, Paul, thanks for coming by. Appreciate it.
Paul:	Thanks.

Paul Lappas of Intermix - A Single Dashboard to Monitor Mission-Critical Data Flows

Explore podcasts from these experts