Rotem Tamir, Ariga | KubeCon + CloudNativeCon North America 2023

This is Textron tv. Hi everyone. We're back here.

Live in Chicago at CubeCon day two afternoon. I guess there must be, uh, a low in the sessions 'cause I see more people walking around the floor. Um, but it's good.

I, I hope by the way, usually, you know, at home we shoot at, at, for those at home we usually shoot into our background, which has our logos and everything. But we tried something a little different this year. We're shooting out into the crowd to give you a sense of, of what's going on here at the show and interested in your feedback.

If you want to, uh, go on on, it's not Twitter anymore, what X or LinkedIn or Facebook or Whatever it is right now. Right? We know.

Or wherever you go. Let us know what you think. We're interested in your comments.

com. Anyway, our next guest is first time here on Techstrong tv. His name is Ro Rotem.

Tamir. Yes. I got it right.

You got it right. Okay. Em, uh, well why don't you tell them your company and your title.

Go ahead. Sure. So I'm the CTO Co-founder of a company called Arriga.

Uh, we started a company something like three years ago, uh, based in Israel and we work on Atlas. Atlas is an open source project, uh, that lets lets you manage your database schema as code. Okay.

So let's back up a little bit 'cause it was fast for the people at home. The name of the company is Arriga. Arriga.

Yes. But how do you spell that? A-R-I-G-A And the website.

io. io. Yep.

Before we jump into Atlas Sure. And maybe we'll go do Atlas first, but Arriga is sort of a commercial company built around Atlas or Atlas That's correct. Is just one of the It's correct.

Uh, we Ariel uh, my co-founder and I started a arriga three years ago around an open source project that he started when he was working at Facebook Uhhuh. Uh, he open sourced an ORM for go, an entity framework for Go, that's called ENT uhhuh. Uh, it gained a lot of popularity.

It's actually part of the Linux Foundation today, and we started the company around that and, uh, Atlas kind of grew from there Really. Yep. So let's talk about the relationship then between Atlas and ant.

Sure. So ANT is an entity framework for go. It means that if you're a backend developer and you need to access a database, ANT is a very popular and convenient way of, uh, doing that.

Basically you define your data modeling Go and ANT has a very strong co-generation engine that can generate a database access code, GRPC servers, graph GraphQL servers. And one of the needs that every entity, framework, or every ORM has is to somehow manage the database schema. And we had something natively part of an for a while, but then, uh, we got a lot more demand and need from the community.

And that's where Atlas kicked in as an independent project. I love it. Glad we got that straight.

Now, Atlas is an independent project Yes. Not part of CNCF or anything. Correct, correct.

Basically your company is is Mm-Hmm. Maintaining It. Maintaining it, yep.

Got it. And, um, we're gonna come back to Atlas now. Let's turn to Monday here at CubeCon.

Yep. You spoke at Argo Conn, which is one of the zero day conferences. Mm-Hmm.

Absolutely. We actually had Dan Garfield here earlier, who is the main one to Maintainance for Argo. I Actually, my Coss speaker was Kois from Dan's company from Res.

Okay. So go fresh. What did you speak about?

So the topic of our talk was, uh, how to do, how to update databases, the GI ops way. So how to apply, uh, GI ops principles to database schema management. Love it.

And how was it received? Uh, I think it went very well. I think there is a lot of need and interest, uh, in this topic actually.

What, what led us to choose this topic among others is that it's one of the most popular or, uh, recurring questions in the Argo, uh, CD Slack channel on, uh, on the C NCF F Slack, You know, and as Dan shared with us earlier, Argo's actually the third most popular Yep. Project N-C-N-C-F now, right? Yep.

Behind Kubernetes and Tele Open Telemetry. Open Telemetry. Yep.

Excellent. Excellent. Now let's talk about Arriga.

Yeah. What does, what does it do around Atlas and, and, you know, how do you, what's the commercial aspect to it? Okay, so, uh, great question.

So Atlas is, uh, a schema management tool. It allows, it's kind of like a terraform for databases. Uh, managing databases is one of these old problems that has been around for a while.

But when we started working on Atlas, we felt that the, the tool set or the, the tools that were available to us as developer were, most of them were like almost 20 years old. And they spawned at the kind of the dawn of the DevOps revolution. So we felt there was a need to, to kind of update the paradigm used like modern DevOps principles, like infrastructures, code and declarative management and, and do like a native, uh, CI integrations.

And, uh, what we saw was the project caught on, uh, very early. We got a front page of Hacker News much earlier than, uh, we had planned, but we, Atlas was supposed to be, uh, some sidekick project. It wasn't meant to be the main thing we do, but, Uh, you know, God laughs at men's plans.

Right? Exactly. Exactly.

And what we've seen in the past year and a half is, uh, the, the amount of downloads for Atlas grow something like 30%, uh, month over month. So it's been becoming, uh, very, very popular. And what we saw was, uh, companies and organizations that started to adopt it have specific needs that come when, you know, uh, the organization becomes larger, uh, and you need, uh, you need more things.

So this June we, uh, released Atlas Cloud. We announced it, uh, in the media. We got a very good, good reception from our community.

And Atlas Cloud is our, uh, commercial product. We have a team plan, an enterprise plan, like, uh, many open source companies do. Yep.

I love it. I love it. Um, so is this your first cube con here, or No, That's a great question.

Uh, I was in Cube Con once before in 2017. Seattle? In Austin.

Oh, in Austin. Okay. In Austin.

Yeah. And I was just, you know, a, a young tech lead back then. I didn't ever think I was gonna start, Actually.

That was very early. Yeah. Yeah.

I think it, it was, there were like 5,000 people then, and it was sensational that so many people care about Kubernetes. I was a young tech lead there. It was a very, uh, I guess, profound experience for me to, to come to this place and have so many, like-minded people that just love technology and love open source.

And I didn't plan to be a founder back then, but I think it kind of had a, a, it's impact in paving my path. What's your impression of this year? It's become, uh, much bigger.

I must say that as, uh, a founder, you have much less time to visit the talks, and it's more about meeting the people. Tell me about it. More about meeting the people and, and I got to meet some, uh, good friends and good partners and to talk to some, uh, some very interesting companies.

And, uh, so that's what, that's what this coupon's been mostly about for me. I mean, the energy's still here, right? The absolutely the enthusiasm for open source.

Yeah. Yep. So I've been going to every CubeCon since, um, let's see, COVID started 2020.

Yeah. So I've been going everyone since 2018. Seattle was the first one I went to.

And um, what always amazes me is, is the enthusiasm of the, of the community. But, um, you're right, it's grown, it's matured. It's not what I, what I've come to realize is back then I only called it Q con.

Yeah. Now it's cloud native, cloud native part more than Q con. Yeah.

In many ways It's become much bigger. Right. The observability piece, the GI ups of, of Argo and everything like that.

Yeah. Serverless security. Right.

It was not as big as Yeah. As, as, as Kubernetes became the operating system for the enterprise. We, we needed as a community to deal with all of these enterprise problems and all these projects spun up in the first, uh, I think after the first coupon I attended, you know, this, uh, slide on the C NCF F website with all the Yeah.

What a mess. Well, it, back then it had like 16 logos, It wasn't as messy. Right.

And now it has I think like over 200. So it's, it's really amazing what everybody is here is doing. Yeah.

And people come from their, uh, domains of expertise and build standards. So, uh, we can build companies and, and really cool stuff upon. I love it.

Any other impressions of CubeCon this year? Um, nothing that comes to mind. No.

Let's talk a little bit more about the company, if you don't mind then. Sure. For our people at home, they're not here.

Yeah. Right. Because, well, maybe they are, but most of them are not here.

Yeah. Um, how did they engage with the company? Like what's the on-ramp for to work with you guys?

Sure. So, uh, at its core, before you talk about Atlas Cloud or, or our, uh, different integrations, I think that anybody that's managing a a backend application has some sort of database. And this database, especially when you're talking about relational databases, uh, this database has a schema, right?

It's made of, of tables and columns and indices and relationships. And this thing kind of evolves over time. And DevOps, if you think about, is really about managing change.

We want to be able to constantly evolve our application as our, our understanding of the world evolves as the market pushes us to do new features. So, what at its core Atlas is about enabling change to your database in a safe way. I think 20 years ago when the, the current like generation of, of database management tools got created, many organiza organizations still had ADBA, right?

They had a professional that read the whole database manual and they really knew exactly what we're doing. Today, teams are expected to manage their own infrastructure. And so a lot of this expertise has kind of, you know, uh, disappeared.

Yeah. And the, the, the idea for creating a specialized tool, the modern DevOps tool for, for managing this change is to codify all of these decades of expertise of managing databases in a way that kind of, the developer just says, I want my database to look like this. And you let automation figure stuff out how to do it in, in, in a safe way.

So the way people find about us is that they look at for modern, uh, uh, database management tools, a lot of, one of the popular Google searches is Terraform for databases. Yeah, sure. People hope that there is something like that and Well, It does.

And that's what comes out for you. Yep. Alright.

I'm gonna ask you one hard question. Please Do. Have you thought about donating Atlas or giving Atlas to the CNCF and let it become A-C-N-C-F project?

That's an excellent question. I think that open source is an amazing, uh, if you think about it like as, as a human race, a very, very amazing way to collaborate and share ideas. I think there are many modes of operating in an open source company.

And, uh, I think the CNCF works best when your goal is to achieve a new standard in the industry. It's one of the things that we're always deliberating with our investors, but we are not yet convinced that this is, uh, the case. It's hard enough to build a, a successful business on open source software, and I think foundations have a lot of advantage in terms of creating a standard, but it really makes things difficult for building a, a successful business.

So I, I Wish So there, there, there's that tension. Yeah. Between, you know, pure, let's call it pure open source Yeah.

And, and a commercial entity. Mm-Hmm. I've been involved in open source a long time.

A long time. And you know, in my world we've seen different, Uh, Modes of open source first was kind of a, I call it the cathedral and the bazaar. Like remember, you know, the book, right?

Yep. Where it was very, uh, laissez fair. It was very, uh, wild, wild west.

Yeah. You know, we didn't want the man in there and all that. Then we went to like, I call it the big brother phase of open source.

Okay. Where you had projects that were really, they were owned by a single vendor. Mm-Hmm.

And, and that vendor ran the project, you know, theoretically for the good of the community, but really for the good of their business. Yep. And then we went into what I call the foundational era of open source Mm-Hmm.

Where, and, you know, it was never as successful as it was in the foundational era. And one of the reason is because when you go with that model, it allows cooperation, right? Yeah.

Companies that would always be competitors cooperate with the idea that the rising tide will live for boat. Yeah, absolutely. And I think that works for a lot of, Of, I think, I think the problem is, and I I really hope we, we figure out a way as an industry around this because this co-opetition is really beautiful and look at all this amazing engineering that people are doing.

Yeah. The problem is when the, when competition isn't fair, it's, it's between David and Goliath. Right.

And, you know, I don't want to call out specific companies, but, uh, not all of them have been equally contributing to open source and, you know, some of them have been Some think writing a, like everything else in life Yeah. Something writing a check is enough. Yeah.

Others think, look, I contribute a lot of code. I shouldn't have to write a check. Yeah.

I don't have that kind of money. Mm-Hmm. Um, and others are just parasites, quite frankly.

Right. Um, I Think I, I think there's the contribution aspects that need to be fair. And, and I think vendors today are more mindful of that and are contributing, uh, employee hours and or just, just money.

But I think, uh, as a startup, it's really hard for me, like as I'm an engineer, right? What do I know about, uh, building a sales organization? I'm learning this right.

But I'm, uh, competing, if I'm competing against a, a 30, 40 year old American company that has a self a, a really well-oiled machine, I think that's where maybe competition is, is, is not like, uh, it's not a level playing Field on any field field. Yeah. So I, and I think though, but that's something that they, uh, the Linux Foundation and, and I don't mean to call it the Linux Foundation, there are other foundations.

Yeah. I think that's something that the foundation has to be the referee. Yeah.

Right. They've gotta try to even that out. 'cause it's not that, that, that's the whole point of it, is that everybody, you know, sort of has a seat at the table.

Yep. And, and granted look, bigger companies Right. Bigger paid, bigger checks and everything else, it's, you know, it's capitalism.

Yep. But it, it, it's absolutely a thing. But I also think that an open source project has to be of a certain maturity Mm-Hmm.

For it to take advantage of, of what a Lennox Foundation has to offer. Agree. Right.

I think it needs, and I don't mean it in the wrong way, but it needs an incubation period before it incubates it here. Yeah. To get sort of a, a critical mass.

Makes sense. I agree. I agree.

I think foundations give a lot of structure to open source. Yeah. And that structure is, is good because, you know, it, it, it provides guardrails, but it's also a, a overhead Right.

That we as, as a young company not always can afford, uh, to pay. Agreed. Well, Tim, thank You so much.

Thanks a much lot, man. Thanks for having me. What's the website again?

Uh, the, I'll give the website for Atlas Atlas. io. io.

Go check that out. We are live in Houston. Houston.

We're live in Chicago. Chicago for CubeCon. We'll be back in a minute with our next guest.