 From theCUBE Studios in Palo Alto in Boston, connecting with thought leaders around the globe, these are Cloud Native Insights. Hi, I'm Stu Miniman, the host of Cloud Native Insights. When we talk about Cloud Native, we're talking about how customers can take advantage of the innovation and agility that's out there in the clouds. One of the undercurrents, you know, not so hidden if you've been watching the program so far. We've talked a bit about serverless. So it's something that's helping remove the friction, allow developers to take advantage of technology and definitely move really fast. So I'm really happy to welcome to the program coming from Fauna. First of all, I have the CTO and co-founder who is Evan Weaver. And also joining him is the new CEO, Eric Berg, that's had both from Fauna talking serverless, talking data as an API, and talking the modern database. So first of all, thank you both for joining us. Thanks for having us Stu. Okay, good to be here. All right, so Evan, we're going to start with you. I love, you know, talking to founders always, if you could take us back a little bit. Fauna as a project first before it was a company. You of course were an early employee at Twitter. So if you could just bring us back a little bit, you know, what created the Fauna project and, you know, bring us through a brief history if you would. So I was employee 15 at Twitter. I joined in 2008. And, you know, I had a database background. I was sort of a performance analyst and worked on Ruby on Railsites at CNET Networks with a team that went on to found GitHub actually. And I went to Twitter because I wanted Twitter the product to stay alive and for no, you know, no greater ambition than that. And I ended up running the backend engineering team there and building out all the distributed storage for the core business objects, tweets, timelines, the social graph, image storage, the cash, that kind of thing. And, you know, this was early in the cloud era. APIs were new and weird. You know, you couldn't get Amazon EC2 off the shelf easily. We were racking hardware in a co-location center and there were no databases or platforms for data of any kind. They really let us, the Twitter engineering team, focus on building the product. And we did a lot of open source work there, some of which has influenced Fauna. Originally Twitter's open source, you know, was hosted on the Fauna GitHub account, which predated Twitter, like you mentioned. And I was there for four years, built out the team, basically scaled the site, especially scaled the twitter.com API. And, you know, we just never found a platform which was suitable for what we were trying to accomplish. Like a lot of what Twitter did was itself a platform. You know, we had developers all over the world using the Twitter API to interact with tweets. And we were frustrated that we basically had to become specialists in data systems because there wasn't a data API, we could just build the product on. And ultimately, that data API that we wished we had is now Fauna. Well, it's a story we've loved hearing and it's a fascinating one is that the marketplace wasn't doing what we needed. Often open source is a piece of that. How do we scale that out? How do we build that? Realize that the problem that you have is what others have? And, you know, hey, maybe there's a company. So, could you give us that transition? You know, Fauna as a product, as a company, where was it understood that, hey, there's a lot of other people that can take advantage from some of the same tools that you needed before? I mean, we saw it, you know, we saw it in the developers working with the Twitter platform. You know, we weren't like, you know, your traditional database experiences either manage cloud or on-prem, you have to administrate the machine and you're responsible for its security and its availability and its location and backups and all that kind of thing. You know, people building against Twitter's API weren't doing that, they were just using the web interface that we provided to them. It was all our responsibility as a platform provider. We saw lots of successful companies being built on the API, but obviously it was limited specifically to interacting with tweets. And we also saw peers from Twitter who went on to found companies, other people we knew in the startup scene, you know, struggling to just get something out the door because they did do all this, you know, undifferentiated heavy lifting, which didn't contribute to their product at all. You know, if they did succeed, then they struggled with scalability problems and security problems and that kind of thing. And, you know, I think it's been a drag on the market overall, where essentially, you know, cloud services were more or less built for the enterprise, for mature and, you know, mid-market and enterprise companies that already had resources to put behind these things. Then there wasn't sort of the cloud equivalent of the web, where, you know, individuals, people with fewer resources, people starting new projects, you know, people doing more speculative work, which is what we originally in Jack was doing at Twitter, you just get going and build dynamic web applications. So I think, you know, the move to cloud kind of left this gap, which ultimately is starting to be filled with serverless in particular, that, you know, we sort of backtracked from the productivity of the 90s with the lamp era, you can do everything on a single machine, nobody bothered you, you didn't have to pay anyone, you know, just, you know, RPM installing, you're good to go. To this Kubernetes containers, cloud, multi-site, multi-region world, where it's just too hard to get a basic product out the door. And now serverless has sort of brought that around full circle, and we see people building those products again, because the tools are finally matured. Well, Evan, I really appreciate you helping set the table. I think you clearly articulated some of the big challenges we're seeing in the industry right now. Eric, I want to bring you into the conversation. So you relatively recently brought in a CEO, came from Akta, a company that is also doing quite well. So give us if you could really the business opportunity here. You know, serverless is, you know, not exactly the most mature market. There's a lot of interest and excitement. We've been tracking it for years and see some good growth, but, you know, what brought you in and you know, what do you see as that big opportunity? Yeah, absolutely. So, you know, the first thing I'll comment on is what, you know, when I was looking for my next opportunity, what was really important is to, you know, I think you can build some of the most interesting businesses and companies when there are significant technological shifts happening, you know, Akta, which you mentioned, took advantage of the fact of, you know, SaaS applications really being adopted by enterprise, which back in 2009, you know, wasn't exactly a known thing. And similarly, when I look at Fauna, you know, the move that I haven't talked about, which is really the maturation of serverless and therefore that as an underpinning for a new type of applications is really just starting to take hold. And so then there creates opportunities then for, you know, a variety of different, you know, people in that stack to build interesting businesses. And obviously the databases is, you know, is an incredibly important part of that. And the other thing I mentioned is that, you know, a lot of people don't know this, but there's a very good chunk of Akta's business, which is, you know, what they call their customer identity business, which is basically, you know, servicing up identity as a set of APIs that people can integrate into their applications. And you see a lot of enterprises using this as a part of their digital transformation effort. And so I was very familiar with that model and how prevalent, how much investment, how much pain was out there for customers as they were, you know, every company becoming a software company and needing to, you know, rethink their business and build applications. And so you put those two trends together and you just see that, you know, serverless is going to be able to, you know, meet the needs of a lot of those companies. And, you know, as Evan mentioned, you know, databases in general and traditionally have come with a lot of complexity from an operational perspective. And so when you look at the technology and some of the problems that Fonda has solved in terms of really removing all of that operational burden when it comes to, you know, starting with and scaling a database, not only, you know, locally, but globally, it's sort of a no-brainer. You know, everybody would love to have a database that scales, that is reliable, is secure, and that they don't have to manage. Yeah, Eric, just one follow-up question for you. I think back a few years ago, you talked to companies and it's like, okay, yeah, database is the center of my business. It's a big expense. I have a team that works on it. There's been so much change in the database market that most customers I talk to is I have lots of solutions out there. I'm using Mongo, I've got Snowflake, Amazon has flavors of things I'm looking at. You know, Snowflake just filed for their IPO, so we need the growth in the space. So maybe if you could just, you know, obviously serverless is a differentiation. There's a couple of solutions out there, like from Amazon, whether Aurora serverless solution, but how does Fauna, you know, look to differentiate? Could you give us a little bit of kind of compare to the market out there? Sure, yeah. So at the high level, just to clarify, you know, at the super high level for databases, there tends to be, you know, two types, operational databases, and then data warehouse, you know, which Snowflake is an example of data warehouse, and as you probably know, the former CEO of Snowflake is actually a chairman of Fauna, so Bob Muglia. So we have a lot of good insight into that business. But, you know, Fauna is very much on the operational database side, so the other half of that market, if you will, so really focused on being the core operational store for your application. And, you know, I think, Evan mentioned it a little bit, there's been, you know, a lot of the transformation that's happened if we rewind all the way back to the early 90s, when it was Oracle and the Microsoft SQL server were kind of the big players there. And then as those architectures basically hit limits when it came to applications moving to the web, you have this whole rise in a lot of different NoSQL solutions. But those solutions sort of gave up on some of the promises of a relational database in order to achieve some of the ability to scale and the performance required at the web. But we required then a little bit more sophistication, intelligence in order to be able to basically create logic in your application that could make up for the fact that those databases didn't actually deliver on the promises of traditional relational databases. And so, Enter FADA, and it's really sort of a combination of those two things, which is providing the trust, the security or reliability of a traditional relational database, but offering it as a serverless as we talked about at the scale that you need it for a web application. And so it's a very interesting combination of those capabilities that we think, as everyone was talking about, allows people who don't have large DevOps teams or very sophisticated developers who can code around some of the limitations of these other databases to really be able to use a database for what they're looking for. What I write to it is what I'm going to read from it and that we maintain that commitment and make that super easy. Yeah, it's important to know that part of the reason the operational database, the database for mission critical business data has remained a cost center is because the conventional wisdom was that something like FONA was impossible to build. People said, you can't, you literally cannot in information science create a global API for data, which is transactional and consistent and suitable for relying on for mission critical, user login, banking payments, user generated content, social graphs, internal IT data, anything that's irreplaceable. People said, there can be no general service that can do this. Ubiquitously at global internet scale, you have to do it specifically. So it's sort of like, we had no power company. Instead, you could call up Amazon, they drive a truck with a generator to your house and hook you up and you're like, right on. I didn't have to install the generator myself, but it's not a good experience. It's still a pain in the neck. It's still specific to the location you're at. It's not getting utility computing from the cloud the way that it's been a dream for many decades that we get all our services through brokers and APIs and the web. And it's finally real with serverless. So I want to emphasize that the font of technology is new and novel and based on an inspired by our experience at Twitter and also academic research with some of our advisors like Dr. Daniel Abadi. That's one of the things that attracted the snowflake chairman to our company that we'd solve groundbreaking problems and information science in the cloud, just the way snowflake had. Yeah, well, yeah, whatever. Yeah, please go on, Eric. Yeah, I was just going to add one thing to that, which is in addition, I think when you think about font and you mentioned MongoDB, I think they're one of the great examples of database companies over the last decade who's been able to build a standalone business. And if you look at it from a business model perspective, the thing that was really successful for them is they didn't go in to try to necessarily rip and replace big database migrations. They started involving with a new class of developers and new applications that were being developed and then rode that obviously into sort of a land and expand model into enterprises over time. And so when you think about Fata from a business and a value proposition, it's harnessing the technological innovation that Evan talked about. And then combining this with a product led bottoms of developer first business motion that kind of rides this technological shift into you creating a presence in the database market over time. Well, Evan, I just want to go back to that. It's impossible comment that you made. A lot of people that they learn about a technology and they feel that that's the way the technology works. Serverless is obviously often misunderstood from the name itself to, we had a conversation with Andy Jassy, the CEO of AWS a couple of years ago. And he said, if I could rebuild AWS from the ground up today, it would be using all serverless. That doesn't mean only Lambda, but they're rebuilding a lot of their pieces underneath it. So I look at the container world and we're only starting the last year or so talking about people using databases with Kubernetes and containers. So what is it that allows you to be able to have, as you said, there's the consistency. So we're talking about acid there, not worry about things like cold starts, which are a thing lots of people are concerned about when it comes to serverless. Help us understand a little bit that, what you do and the underlying technologies that you leverage. Yeah, databases are always the last to evolve because they're the riskiest to change and the hardest to build. And basically through the cloud era, we've done this lift and shift of existing on-premises solutions, especially with databases, into cloud machines. But it's still the metaphor of the physical computer, which is the overriding unit of granularity, mental concept, everything. Like you mentioned containers, like we had machines, then we had VMs, now we have containers, it's still a computer. And the database goes in that one computer in one spot and it sits there and you gotta talk to it wherever that is in the world, no matter how far away it is from you. And people said, well, the relational database is great. You can use locks within a single machine to make sure that you're not conflicting your data. When you update it, you're gonna have transactionality, you're gonna have serializability. But what do you do if you wanna make that experience highly available at global scale? And we went through a series of evolutions as an industry from initially the on-prem RDBMS to things like Google's percolator scheme, which essentially scales that up to data center scale and puts different parts of the traditional database on different physical machines on low latency links, but otherwise doesn't change the consistency properties. Then to things like Google Spanner, which rely on synchronized atomic clocks to guarantee consistency. Well, not everyone has synchronized atomic clocks just lying around. And there are also issues with noisy neighbors and tendency and things because you have to make sure that you can always read the clock in a consistent amount of time, not just have the time accurate in the first place. And Fauna is based on an inspired and evolved from an algorithm called Calvin, which came out of a body's lab at Yale. What Calvin does is invert the traditional database relationship and say instead of doing a bunch of work on the disk and then figuring out which transactions won by seeing what time it is, we will create a global predetermined order of transactions, which is arbitrary, by journaling them and replicating them. And then we will use that to essentially derive the time from the transactions, which have already been committed to disk. And then once we know the order, we can say which ones won and didn't win and which happened before, happened after and present the appearance of consistency to all possible observers. And when this paper came out, it came out about a decade ago now, I think. It was very opaque. There's a lot of kind of hand-waving exercises left to the reader, some scary statements about how it wasn't suitable for things that in particular SQL requires. We met my co-founder and I, Matt is Fauna's chief architect. He worked on my team at Twitter on one of the database groups. We were building Fauna, we were doing our market discovery or prototyping and we knew we needed to be a global API. We knew we needed low latency, high performance at global scale. We looked at Spanner and Spanner couldn't do it, but we found that this paper proposed a way that could and we can see based in our experience at Twitter that you could overcome all these obstacles, which had made the paper overall be neglected by industry. And it took us quite a while to implement it at industrial quality and scale, to qualify it with analysts and others proved to the world that it was real. And Eric mentioned Mongo, we did a lot of work with Cassandra as well at Twitter and we're early in the Cassandra community. Like I wrote the first tutorial for Cassandra for data stacks was founded. These vendors were telling people that you could not have transactionality and scale at the same time that it was literally impossible. And then we had this incrementalism like things with Spanner. And it wasn't until Fauna that anyone had proved to the world that that just wasn't true, that there was more marketing around their failure to solve the information science problem than something fundamental. Eric, I'm wondering if you're able to share just to order a magnitude, how many customers you have out there from a partnership standpoint, would like to understand a little bit how you work or fit into the public cloud ecosystems out there. I noticed that Alphabet's General Venture Fund was one of the contributors to the last raise. And obviously there's some underlying Google technology there. So if you could just customers and ecosystem. Yeah, yeah. So as I mentioned, we've had a very aggressive product-led developer go to market. And so we have tens of thousands of people now on the service using Fauna at different levels. And now we're focused on, how do we continue to build that momentum? Again, going back to the model of focused on a developer-led model, really what we're focused on there is taking everything that Evan just talked about, which is real and very differentiated in terms of the real core tech in the backend. And then combining that with a developer experience that makes it extremely easy to use. And really we think that's the magic in terms of what Fauna's bringing. So we've got tens of thousands of users and we've got more signing up every day, coming to the service. We have an aggressive free plan there and then they can migrate up to higher paying plans as they consume over time. And the ecosystem, we're aggressively playing in the broader serverless ecosystem. So whether it's, what we're looking at is as Evan mentioned, sometimes the database is the last thing to change. It's also not necessarily the first thing that a developer starts from when they think about building their application or their website. And so we're plugging into the larger serverless ecosystem where people are making their choices about potentially their compute platform or maybe a development platform. Like I know you've talked to the folks over at Jamstack and sorry, at Netlify and Purcell, who are big in the Jamstack community and providing really great workflows for a new web and application developers on these platforms. And then at the compute layer obviously, Amazon, Google, Microsoft all have a serverless compute solution. Cloudflare is doing some really interesting things out at the edge. And so there's a variety of people up and down that stack, if you will, when people are thinking about this new generation of applications that we're plugging into to make sure that the fauna is the default database choice. Wonderful. Last question, Evan, if I could, I love what I got somebody with your background, talk about just so many different technologies, maturing. Give us a little bit as to some of the challenges you see serverless ecosystem, what's being attacked? What do we still need to work on? I mean, serverless is in the same place that the lamp was in the early 90s. We have the open serverless ecosystem with the Jamstack players that Eric mentioned. We have closed proprietary ecosystems like the AWS stack or the Google Firebase stack. To your point, Google has also invested in us, so they're placing their bets widely. But it's the same kind of criticism that Lamp, that Linux, Apache, MySQL, PHP, Pearl got. It's not mature, it's a toy. No one will ever use this for real business. We can't switch from DB2 or Mumps to MySQL. No one is doing that. The movement, the momentum in serverless is real and the challenge now is for all the vendors in collaboration with the community of developers to mature the tools as the products and applications being built on the new, more productive stack also mature. So we have to keep ahead of our audience and make sure we start delivering, and this is partly why Eric is here, that those mid-market and ultimately enterprise requirements so that businesses built on top of Fonda today can grow like Twitter did from small to giant. Yeah, I'd add on to that. This is reminiscent for me of back in 2009 at Okta, we were one of the early ISVs that built on and relied 100% on AWS. At that time, there was still, it was very commonplace for people racking and stacking their own boxes and using Kolo, and we used to have conversations about it. I wonder how long it's gonna be before we exceed the cost of this AWS thing and we go in and run our own data centers. And that would be laughable to even consider today, right? No one would ever even think about that. And I think serverless is in a similar situation where the consumption model is very attractive to get started, some people sitting there, is it gonna be too expensive as I scale? And as Evan mentioned, when we think about a few fast forward kind of with the innovation that we can anticipate both technologically and economically, it's just gonna be the default model that people are gonna wonder why they used to spend all these time managing these machines if they don't have to. Evan and Eric, thank you so much. It's great to hear the progress that you've made and big supporters of the serverless ecosystem, so excited to watch the progress there. Thanks so much. Thanks, James. All right, and I'm Stu Miniman. Stay tuned. Every week we are putting out the cloud native insights. Appreciate it. Thank you for watching.