Investing in the next frontier of open source with Bryan Offutt from Index Ventures

The below is a full (unedited) transcript of a Youtube session / podcasting episode I recorded with Bryan Offutt from Index Ventures in January 2020. You can view the video/listen to the podcast on YoutubeApple PodcastStitcheror wherever you get your podcasts.

Erasmus Elsner 0:00 
Hi, and welcome to another episode of Sand Hill Road, the show where I talk to successful startup founders and their investors, about the companies that they built and invest in. The goal, like always, is to give you a sense of what it’s like to be in their shoes, how their businesses tick, how they got to where they are today, and to learn from their many successes and mistakes. And today, I’m super excited to be joined by Bryan Offutt, who is an software engineer turned investor at Index Ventures. Before joining Index more than a year ago, Bryan has worn many hats. After a computer science degree at Stanford, he joined Palantir first as a software engineer and later as a product manager. He later worked as a product manager at MemSQL, which is a serious D stage startup active in the distributed in-memory SQL database management space. In this session, we will take a deep dive into the commercial open source software segment, where Index Ventures is actively investing, having invested in the likes of Confluent, Cockroach Labs, Elastic and most recently Starburst. But let’s hear it from Bryan himself and let’s jump right in.

Okay, so I’m excited to have you here with me today Bryan. Welcome, Bryan.

Bryan Offutt 1:37 
Thank you. Thank you. Great to be here.

Erasmus Elsner 1:38 
And so before we jump into your work at Index Ventures and take the deep dive into commercial open source segment, I want to take a step back and find out a little bit more about who Bryan is, why don’t you walk us a little bit through your personal journey from engineer turned investor at Index Ventures.

Bryan Offutt 1:55 
Yeah, it’s been a bit of a bit of an odd path kind of taking it way back, I grew up in a horse farm in Maryland, of all places, and my mom grew up on the same farm, my dad grew up five minutes away. And we actually I dial up internet until kind of well into when I was in high school. So a little bit of an odd candidate to be a developer. But I did spend a lot of time on the computer, i a lot of friends were really into video games. And so you know, spent a lot of time around that sort of stuff growing up, was also a quite competitive swimmer growing up and so had the opportunity when I went to college to go to go to Stanford, which was an absolutely incredible experience. And when I got there, I kind of knew I wanted to be an engineer, I grew up really liking math and science and all that good stuff. And not really knowing what type of engineer so I went and I took a bunch of the intro to engineering classes. And I took cs 106, a, the much famed intro course at Stanford. And I was kind of immediately struck and excited by both the instant gratification of computer scientists compared to other engineering disciplines, as well as the creativity that I felt like it unlocked. It was this amazing thing where I could sit down at my computer for a relatively short period of time and make boggle right, which was like so mind blowing for me, because I was terrible at art growing up, so I never had visual art down. And to my other intro engineering classes are just a much more longer process to get to something very tangible and real. And so when I did that, I said, Okay, well, what I want to do when I grow up, which was a question I continued to ask myself for many, many years, I ended up getting an internship at palantir, through a friend of mine, James fosco, who’s still there, I interned there, I absolutely loved it, worked on some infrastructure stuff for three months, went back was an individual contributing engineer working on government intelligence product, then ran a couple of engineering teams on that side of the house was actually where I got kind of my first exposure to open source software volunteer, they’re pretty early adopters of a lot of open source. So we’re using a bunch of the HashiCorp stuff and Kubernetes and Elasticsearch in particular, which when it came to index was quite exciting, eventually became a product manager kind of wanted to move into things that were more conceptual and a little less technical. After volunteer, I went to mem sequel was kind of looking for a smaller company. And so mem sequel is about 100, folks at the time. And there, I worked on everything other than the core database engine, which was really fantastic experience because I got exposure to some of the really, really gnarly technical stuff that happens in building databases. During my time there, I kind of ran the teams from the product perspective that worked on our new monitoring solutions, we we released a new UI product from scratch, which is a great experience as well as deployability. So we had to kind of reimagine the deployability started to work with things like Kubernetes. And so kind of kick started some of the work at memsql for that, and then I ended up here at Index.

Erasmus Elsner 4:47 
Yeah, so let’s talk a little bit about your journey towards Index Ventures and as a Swiss national, I feel deeply related to Index Ventures because it’s not only one of the few European venture firms that made the jump across the Atlantic, but is also it is also a Swiss firm that was started in Geneva, and then moved over to London and is now one of the leading Silicon Valley venture firms. In the commercial open source segment, Index has been quite active in recent years. But tell us a little bit about your journey to join Index Ventures.

Bryan Offutt 5:19 
Yeah, it’s a great question, you know, so when I was leaving MemSQL, I kind of had this period of time where I was thinking through what I wanted to do next. And I had a lot of ideas kind of floating around that I wanted to take my time to really, truly think it through. And venture became one of the things that came to mind it was it was really adventure and going back to school, do product design, actually. So there was a world where I went from engineering to product management to product design. So it’s kind of all stack. But I got introduced to Index in particular through a friend of mine named Michelle Valentine, who’s now at Airtable. And the first person I actually met was Mike, and I was super excited to meet Mike because there’s three products that he invested in, they’re quite different, all of which I really like. The first one is Sonos, which makes speakers obviously, the second one is Blue Bottle, because I’m a big coffee drinker. And the third one was Elastic. And I was like, wow, you know, because my last job at Palantir is kind of to migrate us off of our legacy Lucene implementation into Elasticsearch. So I was like, wow, this is super interesting guy. He’s got this like, wide range of different investments that spans from, you know, my caffeine addiction to my nerdiness. And so I met with Mike, we really hit it off. I think within about a week or so I met the other partners that in San Francisco that is, and I was just really struck by how friendly and collaborative everyone was. And I actually really also enjoyed the European influence that you see in the US office, because that is kind of where our roots are. And I think that manifests itself in a lot of different really beneficial ways. Like, I think we have a beautiful office. And I would attribute that largely to, to the European influence we have coming over as Europeans tend to be a bit more stylish than us Americans. And so it, it seemed like the perfect fit and the next step for my career. And so I think about two, two or so weeks after meeting, Mike, I was I was ready to jump on board.

Erasmus Elsner 7:06 
Great. So let’s jump a little bit into the commercial open source software vertical. And we’ve had a lot of operators on the show here, like Sid Sijbrandij from Gitlab, Joe Doliner from Pachyderm, but also investors such as Joseph Jackson, from OSS Capital, or Patrick Chase from Redpoint. The commercial open source segment itself doesn’t need a lot of introduction, but maybe give us a little bit now that we are in the new year, the state of commercial open source in 2020. Where are we at?

Bryan Offutt 7:37 
Yeah, I mean, I think it’s a super exciting time to be investing in open source or a part of open source in general, you know, I think, when index started investing in open source companies, it was a bit of an out there thesis, right? You know, you had Red Hat, but people kind of viewed it, as you know, that’s the unicorn, how are you going to sell something that’s inherently free. Like, that doesn’t make any sense. And we made a number of investments fairly early on, the first of which I believe was my sequel that ended up being great outcomes. And I think the key insight that people started to have is that you can use open source is basically developer freemium, right, the open core model, where the foundational piece of software is free, and some people will always use it for free. And then you charge for products and services on top of that, over time. And it’s been really exciting as a developer, to see how that’s exploded over the past five or so years in particular. And now, you know, if you’re an infrastructure company, it’s almost the default mode of operating to be an open source company. And I think that the tooling that that has come along over that period of time, it’s only made open source collaboration easier, more efficient, and faster. And so I think you’re going to see it continue to explode as a category move from software infrastructure into kind of other spaces within the broader tech world.

Erasmus Elsner 8:55 
So let’s take a deep dive into a bit of a more technical segment of the vertical that you’ve brought up with respect to machine learning. You mentioned the open sourcing of pre-trained machine learning models, we had Pachyderm here, which is version control for large datasets, which plays in the training data, data lineage side of things. So what you’re talking about here is to open source the the training models themselves. And in particular, you mentioned the BERT models, the bi-directional encoder representations for Transformers models, which sounds highly technical. So maybe let’s take a step back, talk a little bit about what this BERT models are all about, and how you see open source moving into this specific segment.

Bryan Offutt 9:44 
Yeah, so I think there’s kind of two things that are happening in the machine learning world as it pertains to open source, right. The first is kind of the infrastructure pieces for developing models, deploying models, hosting models and monitoring models, right. And there’s a lot of pieces parallels between software in that situation. But instead of building an app or building a piece of software, you’re building a model or a piece of software that has a model as a core component. And so some of the tooling from the software world directly jumps over the machine learning type of world, some of it doesn’t. So I think you’re going to see stuff like data version control, you’re already seeing monitoring tools, deployability tools that are specifically tailor made for machine learning. And I think that’s a really interesting investment category, particularly in the context of open source. The second thing that I think is really interesting is how it you mentioned, which is the open sourcing of some of these really big neural networks. And for bird, for example, for Excel, from Google, Excel net, for Facebook, GPT, two from open AI, there’s these big natural language processing models in particular, that are being developed at kind of leading research institutions who have the, the brains and the infrastructure to train these gigantic models. And then what they do is they take this model that’s been trained for hours and hours and hours and hours and hours and cost a tremendous amount of money to train, they open source the basis of it, and then someone can take it off the shelf and with some fine tuning, and you know, a day or so of training, tailor fit it to their use case. And what I think is really exciting about this is what it will unlock within existing applications with regards to developers being able to pick up this really crazy research grade AI basically, and use it anywhere that they want in their app, right. And language in particular, over the past year and a half or so is really exploded in terms of the advancements that we’ve seen around these sorts of models, sort of similar to what you saw on the 2010 with vision, right, Alex net, and, and all that kind of stuff. And so I think it’s a really exciting opportunity for the big research institutions to really provide and democratize access to some of these crazy machine learning models. Whereas they could kind of keep them for themselves. But instead they share them with the world, which I think is really in the spirit of open source. And I think also in the spirit of research, right? If you think about where these these models are coming from, it’s largely folks who are doing research and research. So it’s been about I’m going to write a paper and share with my peers. So I’m really excited to see what that will unlock in the commercial space, whether as standalone companies built around this sort of stuff, or as enabling features within existing products that previously would have been really difficult to do and a super, you know, powerful, meaningful way.

Erasmus Elsner 12:27 
And it’s super interesting, and how far out you see this moving into the commercial space, especially these off the shelf pre train models. What’s your timeframe there?

Bryan Offutt 12:36 
Yeah, I think it’s I think it’s now you know, there’s a number of companies that I’ve met where their their end goal is not necessarily to be an NLP company. But there are aspects of NLP to what they’re doing, whether it’s like search or sentiment analysis for support tickets, or automated question answering for things of those along those lines. And they’re largely using, you know, Bert to help them do that set up. So I think you’re starting to see it kind of happen.

Erasmus Elsner 13:03 
And it’s already super interesting. Another interesting frontier, in the open source software segment that you mentioned, is open source design. And when I had Sid Sijbrandij here from Gitlab, he talked about how open source or version control move into the editing of video, Ashton Kutcher is an investor in Gitlab and with him we talked about, if you have a movie, right now you get the final edit, what if you can get all the individual camera feed, and all the editing information and all the audio tracks, and you could remix it yourself so that you could have several copies of the same movie. It could be edited by the crowd, basically. But I think what you’re talking about here is that you see open source moving into the design application layer. There was a Kickstarter campaign recently, where they tried to have an open source Linux based design tools such as Figma, or Sketch. I think the Kickstarter campaign didn’t really take off. Am I getting this right? Or what’s your take on this open source design segment,

Bryan Offutt 14:05 
The thing I find really interesting about tools like Figma, is I think that they will start to unlock the ability to do large scale, widely distributed design, in a way that has been very difficult to impossible until fairly recently, until these tools came about. And if you think about open source infrastructure software, the reason why that’s, you know, Richard Stallman would kill me, because open source has obviously been around for a very, very long time. But the reason why commercial open source and open source in general exploded so much over the past 10 years or so, is because of tooling, right? GitHub, Slack, zoom, all of these things make it possible to develop software in a distributed manner extremely efficiently. And even when I was writing code that happens within the same building, right? It’s way faster to slack, someone for a quick answer to a question and to get up and go ask them. So what I’m kind of curious to see over time is whether or not you start to see the same thing with design because now Now you have collaborative tooling, and a broader tool chain in general that’s being built out that can help people do the same sort of thing in a design context, just something that historically, I think has been really, really difficult to do, which is why you’ve seen most open source software stay at the infrastructure level, rather than bubble up to the application level. And why you’ve seen most infrastructure companies that eventually sell a monitoring management solution with a UI. That part is almost always closed source, that’s proprietary, right? Because dealing with the conflict resolution and contributions from the community, and all that kind of stuff is still just a little bit more difficult. With regards to application software in open source, that one is a little bit more of a up in the air one for me, because I think that if you think about open source, it’s it’s two things. It’s a distribution mechanism. And it’s a kind of marketing strategy. Right. And distribution mechanism is its developer freemium, and the marketing strategy is it taps into something in developers, which is that they really, really deeply identify with their tools in a meaningful and relatively personal way. It’s why developers put stickers on their laptops. It’s why you know, the vim Emacs classic argument exists, I think that that is something that has to be tapped into for an open source company to really be successful. And infrastructure as a space for that, generally speaking, you have these sorts of arguments, like why would you use that database, this one’s better. Whereas if you’re trying to sell an application, like a productivity tool, for example, or developers might have some affinity, but it’s not quite as visceral and personal, as some of the lower level stuff actually take a lot of this inspiration from our investment in goat and the way that Danny, here next talks about tribalism in the context of street style, right. It used to be music, where people really identify with genre music they like now it’s the identify with the street style brands. sneaker head culture is an example of that. And I think that that sort of tribalism also exists in a very meaningful way, in the open source in the software community.

Erasmus Elsner 17:00 
Absolutely. And I think this is a great segue to talk a little bit more about community, the open source software community can be one of the kindest and helpful community on the internet I think. But on the other hand, if you upset it, it can also turn into an angry mob if you upset it. Thinking about, you know, the day when Microsoft announced the acquisition of GitHub and 1000s of projects were moved instantly, from GitHub over to Gitlab. The question to me always is how can you really cultivate this community? Especially I think if you have multiple projects around one open source core open source projects, as we had in the case of Hadoop and MapReduce, where we had Hortonworks, Cloudera and MapR, I always asked myself, like, who’s contributing the most? And how do these commercial open source companies tailor their contribution? So let’s talk a little bit about about community. What’s your take on that? And how would you advise your investments to cultivate that?

Bryan Offutt 17:58 
Yeah, that point about the can be really kind until, until they’re not as it’s very funny, as someone who grew up listening to a lot of punk rock music, it’s very similar, right, you’re part of the community. And then when you’re perceived as sort of, quote, unquote, selling out, people get pretty upset, I think with regards to who needs to run the community, my opinion would be, it’s got to be the folks who found that the project, right at the end of the day, they are the fathers and mothers of this thing. And they kind of have the moral authority over the project. And so my personal opinion is that they’re the people that should drive the community, or should at least be the figureheads of that community. I think from the commercial perspective, sometimes when you have projects that are split across multiple commercial entities, it can actually be an investor quite problematic, because you basically have three people competing for the same market. But it forced to pick one of those three, I personally would pick a company that had the creators and the founders of the project every time because in general, that’s the people that community tend to rally around, which I think makes a lot of sense, right? It’s like if you could, you know, buy a painting from Salvador Dali, or someone who was, you know, Salvador Dali, his apprentice, and it’s copying his style, you’d go from the original. And so I think that those people tend to be the leaders of that community. And it does take a lot of work. It’s a lot of effort, a lot of work to really foster a community and hit that balance between, especially if you’re a commercial company, between commercial interests and kind of the roots of open source, right, which is kind of an age old, old debate has been going on for forever and ever.

Erasmus Elsner 19:33 
So let’s talk about your most recent investment, public investment in the commercial open source segment, which is company called Starburst. That announced its 22 million funding round led by Index Ventures in December, with Mike Volpi taking a board seat, the company is looking to monetize the open source Presto distributed query engine, which was originally developed within Facebook. And so let’s talk Little bit about something that we’ve increasingly seen open source projects spun out of large technology conglomerates. Other examples of this include the React and GraphQL, which have also come out of Facebook, but also Golang and Kubernetes, which came out of Google, and Kafka, which came out of LinkedIn. And so my personal pieces on this is a little bit that and the reason why we see these open source projects spinning out of these large tech conglomerates, is because some of the problems that they sort of solve or address are problems that are only encountered in at scale technology environments. For example, for Kubernetes, which came out of the internal cluster management system, Google Borg project, that was something that was only necessary, where you really have huge traffic and problems that are not really encountered when you just run a small website. And the same goes for Kafka, which came out of LinkedIn, which really needed this low latency messaging and streaming system. So talk to us a little bit about this open source project spinning out of these Facebook schools and other large tech companies. What’s your take on that?

Bryan Offutt 21:08 
Yeah, I think it happens, kind of, for three core reasons. The first one, you you hit the nail on the head, I totally agree with you, which is these companies, part of what makes them so special. And part of the reason why they’re so successful is because they’re doing something way ahead of the curve. And for the companies, you mentioned, largely that was things with data, right? Whether that’s streaming or pure volume of data, or poor volume of requests, whatever it might be, they’re experiencing things that the market will eventually experience and your downline, right. And I think big data is an example of that, right. And user analytics, I think is another common one these days. And they’re kind of tied together. But people were doing really detailed user analytics, and really leveraging data at these companies in a way that was way different than what anyone else was doing. And eventually everyone goes, Oh, they’re the thought leaders, they’re crushing this, this is useful to my business, I’m going to do that as well. Right. And so I think that’s the first driver. I think the second driver is particularly in the context of open source, most of the stuff that’s come out has been infrastructure that’s been massively successful. databases, streaming services, query engines, things along those lines. And I think the reality of those those tools, especially having worked at a database company, is they take a really, really, really long time to build, right. And not just to build, but to vet and make sure to actually work and all that kind of stuff. So one of the advantages of a project that starts at big companies like it’s been, it works in production, and one of the most intense environments, if not the most intense environment, the world, right. So you basically get a lot of development time and testing time and time to solidify the product in the comfort of kind of a big a big, large organization. And then I think the third thing is there’s an aspect of the team has done it before. And whether it’s the team has done it before, because they own the project at the company, or whether it’s the team has done it before, in the context of they worked at a company where someone built a similar ish product. And they know intimately, the things they would have redone. You know, the decision that got made a year and you realize four years later was like probably not optimal, you know, all of the engineering stuff that you never could have predicted until until you get a little bit further down the road, you get the opportunity to go, Okay, cool. I know and understand all of those things already. And I can go and either rebuild it to solve or rebuild something similar to solve it. Or when I go to sell it to customers, I already know all the things they’re going to run into before they’ve run into them. So I can help set expectations and debug those issues much quicker. So I think the experience you get coming out of those big organizations is also incredibly useful. And I guess the fourth thing is, I think there’s a talent thing too, right? Where those big companies that make these really interesting infrastructure products, as part of the company tend to have great engineering talent. And so when folks leave, they tend to bring their friends who worked on it as well. And so it’s like a natural kind of exodus of engineers, because people already have personal relationships, which is really helpful, I think, in the early days of a founding company.

Erasmus Elsner 24:09 
And so let’s talk a little bit about Starburst and how you sourced the investment and how it’s built around Presto.

Bryan Offutt 24:18 
Yeah, it’s a funny story, how I got connected with Presto, I joined Index, and I was kind of thinking through, you know, as I had left MemSQL, what were things that people were talking about a lot right within the company, and Presto was one of them. And what I found really interesting is both the engineers and the sales people were talking about Presto, which is a good sign, right? Because the sales people were like, I see it crop up places. And the engineering team was like, Oh, this is actually really cool. And anytime you see both of those things, that’s a good size, like solid tech and commercially viable. And so I literally Google presto company, and servers data came up I said, Great, I guess I’ll find a way to To talk to one of the Starburst folks, and I sent Matt Fuller, the VP of engineering, a cold LinkedIn message, which he still has. And I said, Hey, I’m at Index Ventures, we’ve invested in a bunch of open source stuff, I used to be it, MySQL database company, really excited about presto, we’d love to learn more, we ended up getting in touch interacting kind of regularly over a year long period of time. Separately, mike got in touch with the engineers at Facebook, kind of were the creators of Presto at the company, the to join forces. And we were like, great, we’d love to love to be on board. And next thing you know, we’re, we’re all rocking away in Boston, a great story of, you know, people do respond to LinkedIn messages.

Erasmus Elsner 25:41 
Yeah, let’s let’s talk a little bit about what the value proposition of Starburst is. And the way I understand it, they’re adding a number of enterprise centric features. In particular, with respect to security, I think things like roles-based access control, as well as connectors to enterprise systems such as Terradata or Snowflake, what’s the competitive landscape there? What’s the unique insight here that Starburst has?

Bryan Offutt 26:08 
I think it’s exactly what you hit on. You know, when you think about developing enterprise software software, you’re going to sell to enterprises. That is this infrastructure stuff, particularly like databases, or query type stuff, right? The sort of features that the enterprises want, are generally around security, I think when you’re thinking about monetizing open source software in general, which is a question that comes up all the time, with companies, you you give away the stuff that developers love, for free. And then you tend to monetize on the stuff at the DevOps, and like it managers really care, right? So you give away the query engine or the basic version of the database. But if you want high availability, or role based access control, or audit logs, or Kerberos, integration, or connectors to old legacy databases that no individual developers can be playing around on their laptop, you charge for that. And so that’s kind of very much the the game plan of what Starburst is doing from the monetization perspective. And I think it’s a pretty common way that folks go about monetizing open source software. And then on the investment thesis, I think there’s there’s kind of two things that are really exciting about Starburst. And presto, more generally, one is, you know, data is becoming more disparate, not more centralized. And having a query engine that can pull data from where it lives, rather than having to do all sorts of complex ETL is really appealing. And having that layer sit above those databases is great as a developer tool, because if you’re depending on presto, your dependencies on something that presumably is gonna stick around, and you can kind of switch out the pieces underneath as you see fit, if you want to. And then I think the second really interesting resonant use case for me, the hybrid cloud, in particular, if you have a dupe right, and you have a bunch of data in there, but you want to move your data like to say s3, historically, extremely painful operation. And I went through a couple of cloud migrations at palantir. And it’s really painful to migrate data from on prem to the cloud. And so what you can do at presto is you can do interesting things where maybe you start just querying on prem Hadoop with presto, it’s very performance. And then you start moving some data to s3, and you query both. And either you query both for forever, or you go, okay, cool, I need to use the last six months of data. And once we get to the six months where on prem Hadoop is no longer relevant, and that data is stale, we leave it there for record purposes. And we just shift to only querying s3. So I think it really helps with that story of migrating off of legacy on prem systems and into the cloud, which I think is also pretty interesting. And I believe Twitter is using presto, in kind of exactly that manner.

Erasmus Elsner 28:44 
You’ve been at Index Ventures for a little bit more than a year. Now, if I’m correct, talk to us a little bit about your time there, how you sort of built your funnel, how you’ve been involved with existing portfolio companies, and maybe a little bit about your overall experience.

Bryan Offutt 28:59 
It’s been amazing, you know, I think investing is is definitely quite a shift from from working as an operational person. And I think the way I approach my day to day in the investing role in particular, it’s actually more similar to product management that I would have originally expected. And if you think about what you do as a product manager, it’s an exercise in need finding, right? I’m going to ask customers of my product and kind of adjacent products, what they like, what they don’t like, what they wish they had, I’m going to try to figure out what the common threads are. And I’m going to ask the question of why. And I’m going to hopefully outline a plan where I can work with everyone within the company to build out the features that actually address that need. Investing is very much the same, right? It’s a little bit of a higher level, because you’re not operating in, you know, say the database market. You’re you’re thinking about a broader set of enterprise tooling, but you’re doing the same thing you’re going you’re talking to folks who are using these tools are not using these tools in which they add these tools. You’re doing a lot of user research to understand what they like and don’t like. And you’re going ha, I think there’s a really big need in the market for this particular tool that does not exist, and figuring out what those tools are. So I spend a ton, I tend to spend a lot of my time doing that sort of research. I like, I like to read it, I like to read Hacker News, I like to go to developer oriented conferences, and just understand what people are working on these days. And I like to talk to my friends that work in companies, and whatever that role might be, whether it’s sales, or customer success, or engineering, just to understand what’s the most painful part of their day, right? Because usually, that’s an indication that there’s some sort of opportunity to be had from an investing perspective. And then you can use that to drive further research that takes a lot of different forms, whether it’s kind of forum diving that I mentioned, or watching conference videos at 2x feed or whatever it might be now that this commercial open source segment is gaining a lot of traction and attention existing venture capitalists who might not have invested in that segment before.

Erasmus Elsner 31:02 
I assume that the deals in the segment getting more and more competitive.

Bryan Offutt 31:06 
Yeah, I think it’s gotten much more competitive. Because everyone’s seen the success of some of these open source investments that a variety of different firms have made over the past couple of years. And what a great business model, the open core model can really be, I think the advantage that we have, you know, the way we operate as a firm is incredibly collaborative, right. And there’s a lot of nuances and subtleties to running an open source company that Mike in particular, as you mentioned, has a tremendous amount of exposure to but that that knowledge is shared pretty widely throughout everyone that works at index. And everyone is always quite available to lend their expertise, which is one of the great parts of working here, right? Whether that’s Mike for open source, or shardul, for security, or Danny on more of the consumer side. And so I think that the knowledge sharing within the firm around open source, stemming largely from Mike has been really fantastic. And so I think being able to leverage our experience as a firm around open source is a really great way to kind of differentiate, because it is it is a difficult journey, as all companies are for sure.

Erasmus Elsner 32:07 
So Bryan, thank you so much for being with us here today. Where can people find out more about you? Are you on the Twitter or you have a blog? How can people get in touch with you?

Bryan Offutt 32:20 
Yeah, for sure. So I resurrected my Twitter from 2009. It’s my name at BryanOffutt. So I’m available there. And then LinkedIn is probably the best place for content, planning to write much more content this year. And the majority of that will be a combination of LinkedIn, and then Medium, where I think it’s also just my name.

Erasmus Elsner 32:43 
So yeah, thank you so much for this session, and I’m looking forward to following your journey as an investor.

Bryan Offutt 32:49 
Yeah, thank you so much for having me. Super fun.