Ori Rafael - CEO at Upsolver

(Joel Beasley at 00:00:00) Hello, my friends. Today we are talking to Ori, the CEO and co-founder at Upsolver, and we discuss coaching your teams through failure, how Upsolver's technology makes use of data lakehouses to simplify big data projects, and how to approach hiring when scaling rapidly. All of this right here, right now on the Modern CTO podcast. Here we go. This is the Modern CTO podcast.

(Joel Beasley at 00:00:34) Where are you calling in from today?

(Ori at 00:00:36) From Cupertino, California, Bay Area.

(Joel Beasley at 00:00:39) Nice. What is it like there? Is it good weather right now?

(Ori at 00:00:43) It's always good weather here except for a few days. A bit cloudy, but usually the Bay Area really excels on weather.

(Joel Beasley at 00:00:51) When did you move to California?

(Ori at 00:00:54) So a little more than two years ago. Small company, a little bit of customers, and, you know, all in. But it turned out fine.

(Joel Beasley at 00:01:04) Yeah, you have to go out and connect with the customers that you do have. I saw that you got to spend some time as a database engineer in the army at the IDF, right?

(Ori at 00:01:15) Yeah, I did.

(Joel Beasley at 00:01:16) What was that like?

(Ori at 00:01:18) Well, you know, you take a 21-year-old kid that did a computer science degree but hasn't tried a lot more in his life, and you give them an intelligence database with huge scale, you know, compared to them, and a lot of responsibility—more than you would get in a regular company that hires people on a free basis and not people that enlist on a mandatory basis. So you get very, very good experience, and I think that's kind of why the high-tech domain in Israel is so strong, because the army creates a pool of people that experience a lot while still being at a very young age.

(Joel Beasley at 00:02:00) In my personal experience with meeting several people from Israel is they have a high amount of determination and grit and discipline.

(Ori at 00:02:11) Grit, determination, but on discipline? Yeah, I don't know. Maybe personal discipline, but I don't know. People in Israel are known for breaking the rules.

(Joel Beasley at 00:02:25) Oh man, that's fun. Yeah. I was thinking as you were talking, I said, what would I have done if I had access to an intelligence database and I was 21? I would have probably used it to try to date.

(Ori at 00:02:39) Well, there is a list of about three to four things you go to jail for, and doing private stuff on an intelligence database is one of them.

(Joel Beasley at 00:02:50) Thank God I didn't go. Oh man. All right, so one of the things I was really curious about as I was preparing for this interview is I've heard of data warehouses, I've heard of data lakes, but what is a data lakehouse?

(Ori at 00:03:09) I think the data lakehouse is basically the combination of both. So let's kind of step back and see. Let's go all the way back. When I started being a DBA, you know, all I cared was my operating system. So I did everything with Oracle, and people tried to solve all of their problems in Oracle because the data was in Oracle. So it worked well for BI, but then you want to do machine learning or streaming or text search or scale, and it didn't address all the needs.

(Ori at 00:03:41) So from the database perspective, it was still very, very good for performance, and it was still very good for ease of use because I could speak the language of database. I could speak the language of SQL. But eventually I was bound, and it was very expensive, and I was locked in. So then we saw another movement starting, and that was the early days of Hadoop of a data lake. And data lake gave you the benefits that were missing from a database.

(Ori at 00:04:06) Basically you had your storage layer decoupled from the query engine. So you can take data in a lake and you can send it to whatever database that you wanted. You could apply a multi-database strategy, a best-of-breed strategy, and cost was much lower compared to a database. But on the other hand, you lost all the benefits of a database.

(Ori at 00:04:26) So you lost the ease of use, you lost the good performance. So you needed those big data engineers to do the work. So everything was bottlenecked in IT, and data lakes became known for being data swamps. And now if we kind of fast forward, we basically have the same thing but with new platforms.

(Ori at 00:04:44) So we have Snowflake. It's a great data warehouse—I'm using them as an example, but there are others like BigQuery or Azure that offer a decoupled architecture—but you're still locked in. You still have to do everything. If you treat them as your only data store, you still have to do everything in them.

(Ori at 00:05:01) So you still have the same challenges that you have with Oracle. And if you go to a data lake, today you have the Spark-based platform instead of the Hadoop-based platforms, but it's still Hadoop. It's still the Hadoop ecosystem. You still need the big data engineers. You still need people that write Java, write Scala, write Python.

(Ori at 00:05:19) So basically the same problems, but just 10, 15 years forward. And then the concept of a lakehouse emerged. I think the concept of the lakehouse is trying to combine performance and ease of use from databases with the openness and cost model of a data lake. And you see different companies applying it differently, but that's the why. Combining these two benefits, that's the why for the lakehouse.

(Ori at 00:05:45) So now you're going to see different applications, but creating, moving all of cloud data analytics under one thing called the lakehouse that combines both benefits is basically the goal of many companies, I think, in this space. Upsolver is a part of it.

(Joel Beasley at 00:06:01) So what's your primary product or service that you guys do at Upsolver?

(Ori at 00:06:06) We build cloud—you could call it lakehouses or build cloud data lakes—and we connect the data in the lakes to your preferred analytics house. So basically we are creating the combination of a data lake with the data warehouse. Our secret sauce, so let's say superpower, is the ability to do that using SQL that people already know or a visual interface instead of writing code. And you don't need to know all the intricacies and all the implementation details of Spark or Hadoop. So I was a DBA. We founded Upsolver because we wanted to take the modern cloud data lake architectures and make them accessible for the traditional database people.

(Ori at 00:06:50) And I think that's what we do that's unique in the market. We are not the query engine, but we create the store so you can connect any query engine that you want.

(Joel Beasley at 00:07:01) That's pretty neat. How did you meet your co-founder?

(Ori at 00:07:05) In the army. So at the time I was kind of late in my tenure there. So I was head of all data integration, and he was the CTO of a large data science group. And we started to work together, and we actually tried to solve a problem in the area of advertising. You know, advertising has a lot of data, definitely more than you usually put in an Oracle.

(Ori at 00:07:30) And both my partner and I knew databases very well, and we needed to for the first time really work with the data lake. And I really remember the sense of looking at a problem and saying, if this would take me half a day with SQL, I would solve it on my own. And then going and hiring a big data engineer and seeing them working for 30 days and still not delivering the solution that I wanted. So I really—it created the need for me to apply the database language that I know on top of cloud data lakes. And Yoni, my partner being Yoni, is very strong with data infrastructure.

(Ori at 00:08:12) We started building internal tooling, and those tools that we built for ourselves turned into Upsolver.

(Joel Beasley at 00:08:19) And how did you come up with the name Upsolver?

(Ori at 00:08:22) Well, it's solving in a better way. So it's up-solve. That's it.

(Joel Beasley at 00:08:28) That is so good. I love it.

(Ori at 00:08:31) Same car but different technology.

(Joel Beasley at 00:08:33) There you go. I love it. I love it. So what do your customers look like?

(Ori at 00:08:38) So we have a lot of digital native customers because these are customers that want to iterate fast on data, and being able to do something that would take a month in a day is critical for their business. And we also work with a lot of traditional enterprises that don't want to invest a lot of time. Big data engineering is not their core competency. And so the same customer that would buy Informatica would buy Upsolver in the world of data lakes for the exact same reasons.

(Ori at 00:09:08) These are the two main types. It's not that we only go to enterprise or only to small, and most of our customers come to us. So that's how we discovered it.

(Joel Beasley at 00:09:18) That's the best type of business, right?

(Ori at 00:09:21) Yeah, I agree. Did you feel that we need to cover examples? We haven't covered example. I wasn't sure if that's something that you—

(Joel Beasley at 00:09:30) If you have an example, give me an example. Yeah.

(Ori at 00:09:33) So I think I'll give you the following. I'm not sure if you've heard of ironSource. ironSource is a mobile advertising company. They provide SDKs. So sometimes when you see these rewarded videos, there's a good chance it's coming from them.

(Ori at 00:09:48) They're going to at least—what's public is they're going to an IPO this year. Very impressive company with the amount of revenue that they generated. And the scale is crazy, and this is one of the reasons it was featured on the AWS big data blog. And they generate today—they run through Upsolver—five petabytes of data every month.

(Ori at 00:10:13) So that's something I definitely haven't seen even during my intelligence days, definitely a data lake scale. And I met—they decided to move from a data warehouse, cloud data warehouse to a data lake. And they were working on AWS, and they were with the dilemma: should I build a data lake or should I take a product that builds the data lake out of the box?

(Ori at 00:10:36) So classic build versus buy. And the reason they chose us—and in the beginning I was very focused on the fact that we are faster to value. So what the data lake they thought it would take a year to develop, I think in four weeks we were already in production. That's very nice.

(Ori at 00:10:53) But what I found out, and I really liked, is that they defined us as the platform that gave access to data. So they said, we have—let's say I'm giving out—I'm not giving the exact numbers here—but let's say we have an R&D group with 100 people. So they say out of that 100 people, four people can serve the rest of the organization. Every data request needs to be funneled through these four people. So this is a classic IT bottleneck, the classic data silo.

(Ori at 00:11:22) And with Upsolver you have more than 50% of the organization having access to the data, and they can self-serve their own data without going to that specific bottleneck. So we had about 15 times more users compared to building your own data lake with Spark, and we did the project in about one-tenth of the time. And I think that was a very good and impressive success. The scale just makes it more interesting because it's five petabytes per month, millions of events per second, hundreds of instances running at any time. So definitely the scale makes it interesting, but I like the fact that we could take such a digital native organization that works so well with data and still help them.

(Ori at 00:12:07) And I really like that example.

(Joel Beasley at 00:12:10) Dude, that is really cool. That is a lot of data.

(Ori at 00:12:14) Yeah, lots of data.

(Joel Beasley at 00:12:17) So when you saw all that happening, had you already talked with them and partnered with them, or did they just pick you up right off the app store and start using it?

(Ori at 00:12:25) No. They went to our website. It was very early stage. They asked for a demo. Our process is we need to show the users that ease of use isn't just something that we promised in marketing, that it's real.

(Ori at 00:12:38) So after a demo, we let the customers log into the product, connect the real data, and do a POC on any scale. So in a few hours you can have a production scale use case running, and that's what they did. And they started with one use case, and they were in production on that use case after a few weeks. And I think they increased the number of use cases by at least from one to at least 10 different sources that are being used, that are being sent into Upsolver. So that's the growth that we want to see.

(Ori at 00:13:12) We don't want you to really rely on us to come and install the product and, you know, do the implementation on our own. You do the implementation. We just explain how should you model stuff in Upsolver, what's the best way to architect. So we consult, but you do it on your own. Otherwise you're still dependent on the vendor and you are still stuck in the old data lake days instead of being in the new cloud data lake days.

(Joel Beasley at 00:13:36) And what's the most common use case you see across your customers?

(Ori at 00:13:40) The most common use case—I would say that there are two main use cases. One use case, and that's the most common one we have, is that you're doing your analytics on top of your lake. Lake would be S3, for example. Amazon S3 would be your lake. So instead of sending the data to another warehouse, you're going to query directly from the lake.

(Ori at 00:14:03) That packs a lot of complexity because you need to structure the data in a good enough way for performance to make sense, and it's also harder to structure the data because you need to actually optimize the file system. That's something which is hard to do that we've abstracted away, and many of our customers use it. But if all your data is in the data lake, sometimes you want to use another product because you think that product is better to query data from the lake. For example, text search. Text search is a problem that people like to use Elasticsearch or Splunk for.

(Ori at 00:14:35) So you use the lake as a staging area, and then Upsolver is being used to just send the data to the target. So this is usually the combination. We store all your data in the lake. We can make it queryable, or we can send it to the database you already use. So you make the choice.

(Ori at 00:14:51) We don't intervene in your decision of choosing a query engine. You're going to use what you prefer. We're just going to make it easy for you to get the data.

(Joel Beasley at 00:14:59) That's pretty cool. I like it. I really like what you guys are doing.

(Ori at 00:15:03) Cool, thank you. That's the idea of an open lakehouse. That's exactly it. It's not the lock-in of Oracle, but it's still the advantage of a data lake. That's what I see in that term.

(Joel Beasley at 00:15:16) What's the long term—where is Upsolver in 10 years? What's the long-term vision?

(Ori at 00:15:22) Well, this is—I hope this wouldn't sound too philosophical, but I'll try. I think that the world of databases is going to a major change. If you're going to—actually try this. Talk to 10 heads of data and ask them which database technology they're going to use in three years, and they're going to tell you Snowflake and Databricks and Starburst and, you know, kind of all the cool startup names right now in the area of data.

(Ori at 00:15:50) And then ask them what they're going to use in 10 years. And you're going to find out that the answer is S3. So their perception of a database is storage. Storage is such a key piece for databases. So I believe in 10 years, all data would be on cloud object storage.

(Ori at 00:16:07) So everyone would move to the cloud and everything would be stored on cloud object storage. And that's dramatically changed how database works because I think the main thing about the database, you kept going to the database because it held the storage. It owned the storage layer, and now it doesn't own the storage layer. So another platform needs to own that. In the old days, an ETL would write the data to a database.

(Ori at 00:16:31) Now the platform working before the database actually governs the storage, and therefore it must be used as often as you use your database. So I imagine Upsolver to be the platform you use to build and manage your data lake and connect your data to whatever query engine that you want. And I think query engines are going to be much more commodity. So it's not going to be the Oracle, Teradata, SQL Server game. You're going to see tens, if not hundreds of different query engines that are more tailored for the use case.

(Ori at 00:17:04) And how are they going to get the data? That's going to be Upsolver.

(Joel Beasley at 00:17:08) Woah. You just opened up my mind to a whole new world. Have you heard of them storing data in glass or storing data inside of DNA?

(Ori at 00:17:19) Storing data inside of DNA? Not an expert on that field. So maybe in ten years, all data would be stored in that way, and I'm completely wrong.

(Joel Beasley at 00:17:29) Oh yeah. I've got an episode for you. So I interviewed this guy from Catalog, and they store data inside of DNA. And the ratio is a football stadium full of storage servers compressed down to something that can fit in your hand. And it exists today.

(Joel Beasley at 00:17:50) Isn't that crazy?

(Ori at 00:17:52) Yeah.

(Joel Beasley at 00:17:52) Yeah. It's ridiculous.

(Ori at 00:17:54) Storage is the most important thing for a database. The rest, many can do. But any improvement in storage is just changing the physics, just setting the laws of the game. So yeah, that's a game changer.

(Joel Beasley at 00:18:07) So I've got a question for you here. This past week, I was talking to John. He's the founder of a company called Blackpoint. Right? And they do nation state cybersecurity, really cool guy. Worked for the NSA, all that type of stuff. And I was curious, you know, when you have these data warehouses or these data lakes, and you're working with these multiple providers and stakeholders, you know, how do you think about security?

(Ori at 00:18:39) Great question. So I'm going to tie it to how we decided to architect Upsolver. I think the Cloud really changed the name of the game because in the old days you would store data on premise. It was secured. We can start arguing about object level permit, column level permission, and the whole level permit. Eventually, people knew how to do that and the data would still reside in house. So you're kind of a bit more concerned from your in house employee than you're concerned about the external attackers. In the age of the cloud, you want to get the ease of use of the cloud. So now you need to start trusting others with your data, and that's a problem. And then add some GDPR and CCPA seasoning on top of that just to make things worse.

(Ori at 00:19:30) The way we structured Upsolver is that we have a control plane which is a SaaS service. Basically, a user interface where you go to govern Upsolver. But the actual processing that we do, we do inside the customer's VPC. We do it in a way that we can spin up and spin down instances so we can fully manage the platform in their account, but we don't have access to the data itself. So only the customer has that access.

(Ori at 00:20:02) And that creates a new dynamic where you get the ease of use of the cloud, but you still don't have to trust your vendor with your data. And that made compliance much easier. So, for example, Upsolver is still not SOC 2 and HIPAA compliance, but we have customers that have very, very strict data privacy laws. Why? Because they don't need to trust us. So it made our life in selling the product much easier since compliance is just a blocker. There is, let's tell me what you do, but before telling me what you do, tell me if I need to trust you. So that's a major concern when you're selling today.

(Joel Beasley at 00:20:42) That sounds like a huge benefit of what you do. I mean, you're giving them these modern features and you're in a zero trust environment. They don't have to trust you at all. That's awesome.

(Ori at 00:20:52) Yeah. They don't need to trust me. I can't access their data. It's their data.

(Joel Beasley at 00:20:57) I think that's how the future of data is going to be. I think we're going to have our pack of data and we're going to let things in and they're going to perform functions. We're going to see results and we're not going to have to give our data away to everybody.

(Ori at 00:21:11) Yes. And I think that makes sense talking from, let's say, talking not as the CEO of Upsolver as a human. That's what I think should be done. We've seen how people mistreat, misuse and mistreat data, and that's definitely not the future we want to see.

(Joel Beasley at 00:21:28) So I was curious to know, you're a leader. You're building this company. Right? And one of the questions that I get a lot from people that reach out from the podcast is establishing yourself as a leader. So I'm curious to know, how do you establish yourself as a leader?

(Ori at 00:21:46) Hard question. So there is a leader within the company and the leader external to the company. I think for both, you need to build a team of people that can be better than you or are better than you, and not try to micromanage everyone. I think that's very important. And looking in from a side of an investor, if they see a good team and not heroes, then they believe that the company can scale. And within the organization, you're showing that people can become really meaningful in the organization, and that sets a good example for the employees.

(Ori at 00:22:18) I would say the second thing is to lead by example and not do things that I wouldn't expect my employees to do. That's always very important. And I try to write from time to time, go on podcasts, communicate what I think so people would hear it, and I like to also have a debate about it. Definitely, I think my point for improvement for this year is to communicate more. More internally to my team and more externally to the world and what we are doing and why we are doing it.

(Joel Beasley at 00:22:59) Yeah. Because it's really cool. We're going to help you get the word out as much as possible. I think it's very fascinating what you guys are doing. Do you think that being in the army helped prepare you for this? Did you have any leadership lessons from being in the army that helped you for your role leading at Upsolver?

(Ori at 00:23:21) Yeah. I think it was monumental in that sense. Since I think at the age of 23, I was already managing an R&D team, and having ownership on product and project management and, you know, servicing users. You don't get that usually at that age, and that gave me a lot of experience and understanding what and so, eventually, the army is a large enterprise. So understanding what a large enterprise wants. And I think it also helped me. So that's, let's say, the positive side to it.

(Ori at 00:23:48) The negative side to it is that you sometimes have the dynamic of a large enterprise where you have politics and you have other people that want to keep each other happy. And I think it taught me a very important product lesson that eventually, harmony doesn't create good products. You need to be able to manage friction in a strong way. Disagree and commit is one DNA thing that's implemented in Upsolver, and I expect people in Upsolver to state their mind, say what they think even if it's not within their domain, and someone needs to address that and need to address it in a very specific way. And if someone is telling you that what you did is not good, then they're not trying to hurt you. They're trying to get a better product. And if you manage that friction well, you build a good product.

(Ori at 00:24:34) And that's, I think, something that I really took from my time in the enterprise. And one of the reasons I wanted to start a company because what I love the most is I love products, and I like to build good product. And feeling like a mediocre product is something that I really didn't want to do and one of the reasons that we decided to build the company.

(Joel Beasley at 00:25:03) I love that you said that. You said that so well. Someone disagreeing with you isn't attacking you or saying you did something wrong isn't attacking you. It's you're just moving the thing forward. Right?

(Ori at 00:25:15) Exactly. You need to have a lot of patience. Yeah. You need to be experienced and have some confidence and understanding that it's not about you. It's about doing better together.

(Joel Beasley at 00:25:27) Now you've got me curious. What other culture items you have, disagree and commit? What else do you have?

(Ori at 00:25:34) I think that being very attentive and obsessed about customers. You always, I always find that fascinating when you talk to people from AWS, how well that's implemented on a huge company. Our customer, they talk to AWS without giving a specific example and how this serves the specific customer. It's the conversation. It's always from the customer and then track, kind of work your way back from what the customer wants. And I think we have been very attentive to what customers are saying.

(Ori at 00:26:07) You know, it sometimes can be an art on what to listen to and what you don't want to listen to. And sometimes the feedback finds you, kind of puts you in a very uncomfortable place. The truth sometimes can be a little painful or scary, but, eventually, you need to give that feedback a lot of room. And I think that's something that I've seen done well by many people in Upsolver. So I would say maybe that these are the strongest two. So focusing on customers and listening to customers and disagree and commit.

(Joel Beasley at 00:26:41) Yeah. I have noticed that companies have gotten much better at listening to customers even all the way down to when you're working with, like, a Facebook API or LinkedIn API. The steps from five years ago are so different. Five years ago, you would apply for a key and list a description. Now you're doing video walkthroughs of how you're using the product and how you're doing the integrations. Companies are really getting a lot better at listening to their customers. You said it's part of your culture. It's one of the most important two things. But how do you do it? How do you train people to do that?

(Ori at 00:27:18) Well, it's not a one time thing. I think that as the employee that I interview, I talk about DNA. If you'll go to upsolver.com to about us, you'll see that we have our core values published since we were 15 people. As we scale as an organization, we are trying to figure out how to embed that with people with ongoing example. If someone disagrees, let's say someone goes above their manager and comes talking to me, I'm using that for the rest of the team as a good example of good behavior at Upsolver. It's something that I respect, and I'm going to give that person a proper answer.

(Ori at 00:27:53) And if I see someone acting from an ego or a politic or saying that that's not their business, then I'm going to call them out because that's not something to. So if you do that enough time and enough people are getting these examples, then they start implementing them on their own. I was fortunate, by the way, that my core team all, you didn't need to encourage them to disagree. They all disagree. So that really helps that they they did that.

(Joel Beasley at 00:28:31) And then how is it when you're dealing with failure with your teams? Let's say they're trying to do something new, difficult, and then they fail. How do you approach failure within your teams?

(Ori at 00:28:43) Well, we fail all the time. You know? I wake up in the morning and fail twice, and that's kind of what startups do. So there's no judgment. You're allowed to make a bet and find out that if it's the wrong bet. That's kind of the way things work. So I don't think that I've ever been mad at anyone for not succeeding. I wouldn't be happy if someone wouldn't communicate, wouldn't listen, do their own thing without explaining. It's more the behavior around how you manage your bet than if the bet was successful. And that's the way we see it.

(Joel Beasley at 00:29:28) You just made an Instagram quote. That sounded really good. Yeah. It because it's true. I write it down. Yeah. Yeah. It is. It is so important. You're right. It's not that you lose. It's your behavior throughout the entire process. So so what are you learning right now as a leader?

(Ori at 00:29:47) I think that right now, Upsolver is in a pretty aggressive scaling process. So we were 15 people six months ago. We are 30 now, and we're supposed to be 70 by the end of the year. So I'm learning how to hire good managers, good executives that can own their piece of the business. How to, from how to interview to how to test. And for example, you asked me about DNA. The best way to keep your DNA is not to hire people that don't have the right DNA for a startup. So you really want to check that during the interviewing process, especially with an executive.

(Ori at 00:30:13) So before hiring a person, you're going to talk to his previous boss, his subordinates, his peers, their peers. I think that's been very, very important. And for me, learning how to build such a team and how to structure that team and think large organization and not thinking as a team, that's definitely a change, but a welcome change.

(Joel Beasley at 00:30:54) That's so interesting that you say that because in our team meeting this morning, we were discussing the importance of checking references and the value that the references have when making new hires because they will tell you everything.

(Ori at 00:31:09) Yeah. If I ask someone, would you hire that person again? And the answer is no or a slow yes, then it's probably a no.

(Joel Beasley at 00:31:20) So what other, did you ever make any big mistakes when hiring? Or what did you learn from those mistakes?

(Ori at 00:31:28) Yeah. I've made mistakes. Definitely, I won't be specific.

(Joel Beasley at 00:31:32) Come on. No. I'm kidding.

(Ori at 00:31:34) Just to kind of keep the people's privacy. But I've learned that when I feel that it's not that, you can't really change people. So if it doesn't work, someone once told me, I think that's the extreme of it, that if you are wondering whether you should fire someone, you're probably three months too late. So that's taking it to the extreme, but, eventually, we're not here to try and change people's DNA. You can teach the material. You can teach the domain, but you can't really teach a person how to act. You can, if it's, let's say, you can improve an 80 to a 100, but you can't improve a 50 to a 100 as an employer. And eventually, that sinks the rest of the organization.

(Joel Beasley at 00:32:24) It has to be their choice to want to improve on the behavior items. You can't force people to do it. You can't even really incentivize them that much to do it, to change their core DNA. The only thing I found to be successful is just shutting the door on the relationship and then eventually, they get enough doors shut that they have some self reflective experience. Because I think that's what happened to me because I used to not be great. And then I kept coming up against things that wouldn't work for me. And then I realized, okay, I heard this guy talking and he had said the most frustrating thing in life is wanting above average results without being an above average person. I was like, oh, man. I'm not an above average person. So then I went on this, you know, personal journey to improve myself and grow myself and it's you get the results.

(Ori at 00:33:16) I agree. I agree. And I think step number one is awareness. Usually, what blocks awareness is ego. So as long as that person is willing to be aware and improve, usually that will happen and I can probably give a lot of examples of things that I've failed to do and didn't feel like I have accomplished enough. You know, I really like the test. Would I regret it? So let's say tomorrow, I have to shut my startup down. If I wouldn't do this move right now, would I regret it? And that kind of takes you out of your shell and forces you to do something that maybe you feel uncomfortable with. And I think that's kind of the best people I see keep forcing themselves outside of their shells, but that's definitely not everyone.

(Joel Beasley at 00:34:06) Yes. I know exactly what you mean. I know what maximum difficult is. Right? I know that feeling of maximum pain, maximum difficulty, and I don't live there full time, but I am a frequent visitor. Because that's the only way to grow. Otherwise, you just get bored with life. You're like, nothing's really happening. I need some challenging difficult thing and you go for it. Yeah. Where, how do you get inspired? Do you watch YouTube videos? Do you read leadership books? Where do you get inspiration from?

(Ori at 00:34:41) Well, I think that eventually I get inspiration. So I'm a huge management book nerd, but it's been a really long time since I've done that. I used to really like, during my MBA and before and after, I read a lot of management books, so I think I understand the theory. But eventually, when you get down to business, you have a lot of plausible situations where there's no really right answer. There's just the least worst option.

(Ori at 00:35:13) And those are the ones that I like to learn from more. So I think that I learned the most out of talking from other CEOs, sometimes talking to my investors, hearing from their experience. Hearing how people did something specific usually is the most inspiring thing for me.

(Joel Beasley at 00:35:30) And so because your company is hiring, do you guys have—I didn't want to gloss over this. So sorry to change the topic. We were just talking so much about hiring and culture, and I was thinking this would be a good time to tell people where your careers page is or how to go learn more about the company so they could apply.

(Ori at 00:35:47) Yeah. Go to upsolver.com to the career pages. There are dozens of open positions, both leadership and not leadership. So from product to solution architects, to a head of solution architects, to engineers, account executives, partnerships. So really across the board, both go-to-market and engineering.

(Joel Beasley at 00:36:14) And are you in multiple countries? Are you working remote? What is your work environment like?

(Ori at 00:36:20) Yeah. So we are dual headquarters. We have a base in Israel that's more engineering focused, and we have a base in the U.S. that's more go-to-market focused. And of course, there are exceptions. Our executive team is located in the Bay Area, and our employees are spread entirely across the entire U.S.

(Ori at 00:36:44) So all—I think about ten states at the moment. We don't have a physical office now, so we shut it down during COVID. So everyone has been working remotely, and that's not optimal. But what can you do? And I really look forward to having an office again, but it's never going to be the same five days a week office like it used to be. Probably two days a week, maybe three days a week. That's the most we're going to get to in our office.

(Joel Beasley at 00:37:12) The freedom is so great. The ability to schedule everything around the work that you have to get done and just letting that work really connect well to the KPIs that drive the business and create revenue. So it becomes very, very clear, here are the actions that we perform that generate revenue and bring value to the market, and then getting that freedom to structure your day around achieving those goals. I mean, I was not much of a remote work fan for my own businesses because I liked having everybody together. It just gave me a sense of security, I guess. I liked having everybody in the room, but after we clarified everything and really, really focused, it's just become a whole other level. And now we can get the best people from anywhere in the world to come work with us.

(Ori at 00:38:03) Exactly. I don't think we could build a team in the Bay Area. We would just not succeed in it. It takes too long, and you have really good people across the entire country instead of just one state.

(Joel Beasley at 00:38:18) Do you think you're going to stay there for a while?

(Ori at 00:38:21) Yeah. I am going to stay here.

(Joel Beasley at 00:38:23) So you're one of the California people. You love California, not going to leave.

(Ori at 00:38:27) Yeah. I like California. You know, it's expensive. That's one drawback of California, but weather is amazing. Nature is great. You can go from snow to the beach at the same, you know, three-hour ride. The ecosystem is here, so you have a lot of good people you can meet around here. So I think being close to the ecosystem is very important. So I didn't move from Israel to the U.S. and then still be a flight away from the rest of the ecosystem. So I'm going to stay here.

(Joel Beasley at 00:39:02) That is very true. Right now, my wife and I are looking to relocate, and one of the biggest decisioning points for us is we need to be near a city that's populated enough to support her business because she does kitchen remodeling. And so you have to have the right market to do that. Right? And it's very important being near people.

(Joel Beasley at 00:39:27) So I'm excited because I think in a couple years when people are meeting more in person, I'll be outside of a very large city, maybe twenty or thirty minutes outside of a very large city, because right now I'm in a small city. And I'm really looking forward to that because you're exactly right. Being close to the ecosystem, being close—you know, it's one thing if I'm going to wake up and it's going to be six to eight hours of travel time to go have this one meeting to develop a relationship in person versus just living in the city where there's thousands of potential people that could help you grow.

(Ori at 00:40:06) Yeah. Makes total sense. I think the only reason to really be in the suburb is, for me of course, we have kids. And we need—we want more room, and the environment is more friendly comparing to the city, but being close enough to the city is important.

(Joel Beasley at 00:40:25) Yeah. So how many kids do you have?

(Ori at 00:40:28) Two. Small ones.

(Joel Beasley at 00:40:30) Yeah. I call them banshees.

(Ori at 00:40:33) That's a good definition.

(Joel Beasley at 00:40:35) Yeah. I've got a daughter who's almost four and a son who's almost two.

(Ori at 00:40:43) Yeah. We are pretty—it's called Irish twins because they're never in sync, and we have the same. So we have a two-year-old boy and a three-and-a-half-year-old girl. So we are in the same boat.

(Joel Beasley at 00:40:55) That is crazy. Yeah. That is really, really, really close.

(Ori at 00:40:59) Yeah. She's a talker. He's a terrorist.

(Joel Beasley at 00:41:06) I think that's—we have—that's the trend. Right? Talker and a—I'm going to write that down because my wife is going to laugh so hard. Man. Okay.

(Joel Beasley at 00:41:19) Can we talk a little bit about the future of technology?

(Ori at 00:41:22) Sure.

(Joel Beasley at 00:41:23) Okay. So Domino's Pizza—they're calling themselves, I saw an article for it, right? They're calling themselves an AI front runner, and they're doing tons of AI at Domino's, but they're also doing self-driving delivery. Have you heard about this?

(Ori at 00:41:41) No. I haven't.

(Joel Beasley at 00:41:42) Would you order self-driving pizza from Domino's?

(Ori at 00:41:45) Why not?

(Joel Beasley at 00:41:46) Exactly. It seems like the logical next step.

(Ori at 00:41:49) I'm not taking the risk. You know? I'm just getting the pizza. What do I care? It's even nicer. Maybe they'll find the place without calling me and asking me how to find it. You know, the DoorDash calls you keep having when you order food.

(Joel Beasley at 00:42:06) So I guess the car pulls up, and then you'll have to go outside and get it out of the car unless they do last foot drone delivery service of some type.

(Ori at 00:42:17) Yeah. I don't know. The drone would be better. So I need to get out, but you know, going down and taking it from a robot is not too bad. I think that's generally a good thing. Domino's were already amazing. I think that if you order pizza from Domino's, you're going to get your Domino's in a very clear time, and you're going to get some communication on what's going on. You know, of course, the DoorDashes of the world changed that because they're providing a similar experience with visibility and mobile app.

(Ori at 00:42:50) But a few years back, Domino's already had a very good advantage from a logistical standpoint. So taking that to the next step and doing it autonomous is very cool.

(Joel Beasley at 00:43:03) They do. You're right. Domino's has been—I haven't thought about this much, but they've been ahead of the game for a long time. I remember on my flip phone refreshing a page to see the status of my order. So that must have been ten years ago I could see the status of my order on Domino's. Yeah.

(Joel Beasley at 00:43:22) We should have their CTO on.

(Ori at 00:43:25) Advances in logistics. Remember the time that people thought there was eBay and there's Amazon, and Amazon has all of this logistics burden on them. That's probably going to sink them. Some people were saying that, but I think it turned out pretty well for them to have that burden.

(Joel Beasley at 00:43:43) Right? I don't even know. I feel like they're going to open up a shipping company, and then you could just ship stuff. I mean, they have an entire infrastructure for shipping. They have all of these Amazon trucks everywhere. UPS and FedEx, they don't even really deliver the packages anymore. Where I live, they all come in these Amazon vans.

(Ori at 00:44:03) Yeah. I would be surprised if they don't already own a bunch of logistic companies, but I don't really know.

(Joel Beasley at 00:44:09) They have so much cash too. They could probably just buy FedEx or UPS if they wanted.

(Ori at 00:44:15) Yeah.

(Joel Beasley at 00:44:16) Yeah. Probably. But they usually don't—if they don't buy them, they'll just beat them. They're Amazon. That's crazy.

(Joel Beasley at 00:44:25) And I got really inspired by reading about some of their culture items there because when you see a system that gets that big, it really highlights how important the principles are that are the foundation of the system.

(Ori at 00:44:41) Yeah. I agree. And Amazon is a very efficient machine. And we talked about the other day how important it is. Look how many different businesses they own well, and how much consistency you see across their employees.

(Joel Beasley at 00:44:55) So you're hiring. You're growing. It's exciting. What's the thing right now that you're thinking most about? What are you really excited about?

(Ori at 00:45:06) Well, I told you I'm a product guy. So I'm excited about the change that we're going to do, the same change that inspired us to build the company. I think this is the golden age of data. So look at the IPOs, the investment happening, the market. This is the golden age of data. And in the combination of the clouds into it where you can completely re-architect the old products. So making heroes out of the traditional database practitioners, that excites me. Every time I think about the next leap in the product that's going to include more of these practitioners, that's the part I'm most excited about. Of course, you know, it's exciting to scale the company and to scale the team and to hire the right people, but I think, eventually, it's serving the vision.

(Joel Beasley at 00:45:54) And do you partner with those companies that you help?

(Ori at 00:45:57) Yeah. So we have a whole bunch of partnerships. I think maybe our most close partnership is with Amazon, by the way, with AWS. So AWS will often recommend Upsolver. And if you check out Twitter, you see that Jeff Barr and their chief evangelist Jeff Barr just tweeted about Upsolver, and he heard that it's a very good product. And we were featured in their last sales kickoff, and we have a whole bunch of customers that came as a result of referrals from AWS, and we often co-sell with AWS. And our product is accessible to the AWS Marketplace. So we've done a lot of work with them, also starting to do some work with Azure. So cloud providers are something very, very close to what we do because we provide the cloud-native architecture. What we can do now wasn't really possible before you could offer cloud-native solutions.

(Joel Beasley at 00:46:51) That's pretty cool. That's a deeper partnership than most. It's not like you just use them as your servers. You guys are actual partners.

(Ori at 00:46:59) Yeah. I think that because we complement their ecosystem, eventually, Amazon is going to give you—I think they have about 250 services. So you want to build a data lake, you need to use three, four, five, six services depending on your complexity, and that takes time. And eventually, it prohibits adoption, and it provides competitors a chance to take business away from them. And then you pull up Upsolver, and you solve the same thing that you would solve in six months, in a week, and that increases adoption. And it's good both for them and for us. So we are increasing their consumption, and they're happy that we are making our revenues from it as well. And it can be good all around also for the customer that doesn't need to spend six months building something.

(Joel Beasley at 00:47:44) Is there anything that we didn't get to today that you want to get out there into the world?

(Ori at 00:47:50) No. I think we have covered it well.

(Joel Beasley at 00:47:55) Boom. We did it, man. Perfect. 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.

(Joel Beasley at 00:48:08) 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.