Hashicorp. The OSS juggernaut itself dropped its S-1 registration last week revealing financial data and other details we’ve been anticipating for years. The Company was launched by college friends Mitchell Hashimoto and Armon Dadgar (Hashicorp is of course named after Mitchell, a slightly better name than Dadcorp or Garcorp) and has grown from a curious amalgamation of a handful of open source products - to a full suite of 8 core products and additional supporting solutions.
The company now boasts ~100 million downloads TTM (trailing twelve months), 2100+ customers, 36,000 community user group members, and $259 million TTM revenues. Impressive to say the least!
We will write more about their OSS products - but there is too much to cover for this issue.
This week, we sat down with Mark Hinkle, the CEO, and Co-Founder of TriggerMesh, a cloud-native platform to build event-driven applications on Kubernetes. TriggerMesh recently announced they were going open source their platform and Mark gave us the scoop!
Private Markets
ClickHouse, providing real-time SQL analytics, announced their $250M Series B at $2B valuation led by Coatue and Altimeter.
Treasure Data, building the leading enterprise Customer Discovery Platform, announced their $234M funding round led by SoftBank Corp.
Yugabyte, a scalable distributed SQL database, announced their $188M Series C at $1.3B valuation led by Sapphire Ventures.
H2O.ai, building a cloud platform for AI systems development, announced their $100M Series E led by Commonwealth Bank of Australia.
OctoML, accelerating ML deployment, announced their $85M Series C led by Tiger Global Management.
FingerprintJS, building a fraud prevention fingerprinting service for the web, announced their $32M Series B led by Craft Ventures.
QuestDB, making an open-source time-series database, announced their $12M Series A led by 467 Capital and Uncorrelated Ventures.
MindsDB, accelerating ML adoption inside databases, announced their $7.6M Seed led by Walden Catalyst Ventures.
Ethyca, building APIs, tools, and analytics to help organizations adhere to privacy policies, announced $7.5M Series A extension led by Lee Fixel, IA Ventures, LGF, and Table Management.
Activeloop, building the Database for AI, announced their $5M Seed led by 468 Capital and CM Ventures
Umbrel, enabling people to run their own servers from home, announced their $3M Seed led by OSS Capital.
Public Markets
To track the performance of COSS companies, we’ve created an equal-weighted index comprised of public names including: Gitlab, Kaltura, Couchbase, Confluent, MongoDB, Elastic, Rapid7, Fastly and Jfrog.
The COSS Index has been in positive territory over the last twelve weeks and has surged to close the gap to major indices this year.
COSS Index +16%
NASDAQ +24%
S&P 500 +25%
The COSS Index continued to jump above the NASDAQ over the last three years as strong appetite for cloud names continued to drive strength.
COSS Index +194%
NASDAQ +112%
S&P 500 +68%
COSS companies jumped to their highest level ever adding nearly 8 multiple turns as revenue multiples hit the stratosphere at nearly 3x the five-year average. All three indices continue to trade significantly higher than their rolling five-year average.
COSS Index: Current Multiple 26.3x | Five-Year Mean: 9.0x
Emerging Cloud Index: Current Multiple 14.7x | Five-Year Mean: 9.9x
NASDAQ Composite: Current Multiple 4.6x | Five-Year Mean: 3.3x
OSS Newsletter:
What is your background?
Mark:
I was born a poor country boy in South Central Pennsylvania that accidentally got into tech and have had the privilege of being at the forefront of most of the major technology trends over the last 25 years. I worked for the first commercial internet service provider that was national, PSINet, got into virtualization in 2000 when we were doing x86 virtualization for SCO UNIX. That's right before VMware got very big. I did a bunch of systems management at open source startups and had a great result from cloud.com. I was their VP of community. Along the way, I was an amateur journalist and editor-in-chief for Linux World Magazine and Enterprise Open Source Magazine in the early 2000s. I started the Open Source Office at Citrix and then left Citrix to be the VP of marketing at the Linux Foundation. After that, I ended up founding TriggerMesh with a former colleague, Sebastian Goasguen, who I'd met through the Apache CloudStack project where we were both committers so a lot of open source stuff.
OSS Newsletter:
What is TriggerMesh?
Mark:
The background is sort of funny because initially, TriggerMesh was not open source. We were leaving our jobs and we knew we wanted to do something in the cloud. We had been trusted lieutenants at places like cloud.com and Bitnami and Sebastian and I had both worked at Citrix together, working on Apache CloudStack. We had created a website and Sebastian was the author of Kubeless which was the first serverless framework for Kubernetes, and Google had asked him to share what he had learned because he had left his job where he was working on Kubeless, so he was a free agent and we knew we were going to have a company called TriggerMesh. We knew it was around event-driven architecture and we had no other idea, and then at Google Next, they made this big fuss about Sebastian, the co-founder of TriggerMesh. That announcement turned into Sid Sijbrandij, the CEO of GitLab calling us up and saying, "Hey, we hear you're really smart. Can you help us build the serverless publishing platform for GitLab," and that's how we bootstrapped the company. We had to hurry up and incorporate so we could bill them and have revenue in the first three weeks in excess of six figures. It was a crazy, crazy genesis to the company.
OSS Newsletter:
Since that initial genesis moment, what has Triggermesh become?
Mark:
TriggerMesh today is a platform for building event-driven architecture of which serverless is a subset of that, and as of two weeks ago, it's open source, so it's cloud-native, built on Kubernetes, and the idea is that we need to get data places in as close to realtime as possible, so we build infrastructure that takes realtime event streams and connects them to places where the data that's put in motion can be actionable, so you may have a change to an ERP system and you want to make sure that you replicate the unique ID to your mobile app in realtime so that when someone signs up, they don't create two duplicate IDs.
One of the things that's really hard for banks is to gather real-time data from all the different sources. A lot of those sources are mainframe, large systems, and some of those sources now are cloud, so they got to have a substrate that allows them to pull data sources and get them replicated as fast as possible.
OSS Newsletter:
What was the rationale for moving to open source?
Mark:
We were always building an open source team. 18 of the 22 people at TriggerMesh have had a significant role in open source from co-founding Pentaho which is open source BI to launching Mulesoft and Hyperledger as PR and marketing or committing to Elastic. We all had this DNA, but the thing about open source is, you rarely are successful if you just put something out there and hope that people get it, and especially when you are a small team, so we wanted to get it to a point where we felt people could consume and understand our ideas and we could have a meaningful dialogue, so we wanted to get to 1.0. and we wanted to make sure that we had thought through the strategy. There's sort of this strategy of these one-off open source type licenses where it may look like the Apache license but prevents people from offering it as a hosted service, and typically the companies that have been successful started out with an open source license and made that change when they got pressure from other providers once they went public, they got to a certain size.
We just wanted to do open source software in its purest form and figure out a way to make our competitors note our experience and our community, and so that's what our thought process was around it. Once we got to the point where we knew enough about what we wanted to build, what was saleable, what could be open, then we open sourced it. Actually, we had been working on open sourcing it for quite some time, but we needed to have enough capital to hire somebody to do developer relations, to have enough engineers to review pool requests, to help people get installed, to make sure the documentation was good. You can plant the seeds but you still need to nurture them, so we planted the open source seeds and they're starting to grow and we already got our first pool request in the first two weeks. That's when you know you might have something interesting because people are contributing, so we already got contributions, so that was the influence, but we built an open source team knowing that eventually, we'd open source it.
We're also sort of perfectionists in that. We wanted to do it right. My friend and our colleague is the president of Apache Software Foundation and his joke was, yes, you're not doing it right when he was criticizing this commercial open source, and I could hear his voice in the back of my head saying, "Make sure you do it right," and I think we did, or are doing it well.
OSS Newsletter:
What is your advice for building great open source communities?
Mark:
You want to make something bigger than yourself or your organization, so, part of our launch included getting VMware and Cisco to buy in and support the project and that could be in the form of documentation, publicity, distribution as well as core development. I think the number one thing is, making it bigger than yourself, and two, providing a collaborative environment where you don't suffer any, let's just say, shenanigans. I've been involved in projects where the leaders have been toxic or there have been members of the community that have been toxic to others, so I feel like I don't want to be part of a cancel culture but I also don't have any kind of tolerance for folks who are not civil to each other.
OSS Newsletter:
Where do you want to take TriggerMesh?
Mark:
We want to be the de facto integration layer for Kubernetes. We feel like Kubernetes is going to be the fabric for the enterprise, and particularly the cloud, but that's the big, hairy, audacious goal, and then we want to make sure that this technology is something that is much bigger than TriggerMesh. In the short term, we want people to use it. In the long term, we want people to contribute back to it, and for it to become something that is part of VMware's cloud-native stack, which it's already, part of it is in VMware stack, part of Cisco's. I would love to see it on Amazon as a product offering and Google and Azure and Red Hat. So that's the goal - to be a pervasive layer across all of Kubernetes.