Portfolio Companies

The world of enterprise infrastructure has been undergoing a fairly dramatic renaissance over the past several years. Where hardware once dictated datacenter capabilities, intelligent software is now the key component in defining and building the new datacenter. For example, you may have previously bought a server by specifying its hardware characteristics, but you can now define a server through software by way of virtualization. As the datacenter continues its transformation—from a static, hardware-based environment to one that is defined by software—the result will be an agile, cost-effective, enterprise-class infrastructure for next-gen computing.

To date, the major innovations in enterprise infrastructure have occurred in the compute and networking layers of the datacenter. VMware transformed the x86 hardware landscape through server virtualization, and more recently, Nicira (acquired by VMware) redefined networking by creating a software framework for networking. Storage is the last big piece of the datacenter that is ripe for disruption. While there have been recent innovations in new storage technologies like flash, storage architectures remain hardware-defined and have not changed fundamentally over the past 20 years.

Through, the potential of the software-defined datacenter will be fully realized—at last—with software-defined storage networking.

The team at is uniquely qualified to usher in a new world of software for storage. The founders have deep technical backgrounds and have been part of previous transformative trends in storage and virtualization. The three co-founders are infrastructure rock stars. Ramana Jonnala started his career at VERITAS where he helped to define the pre-eminent storage architecture of the past 20 years. Ramana was also instrumental at XenSource, which was acquired by Citrix, where he led a variety of virtualization initiatives. Keir Fraser and Andy Warfield have deep virtualization and storage backgrounds. Keir was one of the inventors of Xen and Andy has produced a mountain of storage code for virtualized environments. The team could not be better suited to take on the new storage networking opportunity.

As someone who has worked with this team for many years, I am thrilled to work with and the superb founding team.

As a former programmer and engineer, I sat in awe as Geoff and Matt presented Meteor at our first meeting. The meeting started normally enough: a brief discussion with the team on their backgrounds, a PowerPoint presentation on what Meteor was up to, a lot of nodding around the room, questions along the way. That all changed when Geoff decided to show us Meteor live. In all of 30 seconds, he hacked up a rich-client and back-end server application. The client was running locally yet synchronizing all of its data to a backend server. Fast, secure and in real-time. What I saw in front of my eyes was magic!

JavaScript has become the number one programming language on GitHub, above Java, C, PHP—everything. The reason for this is that Web clients, such as mobile devices and tablets, are becoming richer and more capable as compute devices as opposed to display devices, and running the application logic locally on the client results in performance, usability and scale for users.

The problem with JavaScript, however, is that it was designed to be a client-side language, leaving all the back-end server implementation to other languages. The result is that cloud-based Web applications take way too long to develop due to the sheer complexity and brittleness of the environment. Without Meteor, everything from security, to multi-tenancy, to latency, to database access requires special APIs and custom development, and developers need to know at least two languages.

The Meteor framework solves all of these problems. Meteor makes real-time application development dramatically faster and more approachable. It gives developers a comprehensive platform for writing Web apps in JavaScript where both client and server code use the same language and API, enabling the same code to be run on both the client and server. The result is real-time, cloud-based Web apps that are scalable, secure and distributed by design.

We see this technology as fundamentally important to the future of the Web. Through this investment in Meteor, as well as our recent investment in GitHub, we at a16z are excited to help developers build the next generation of applications.

The Meteor magic would not be possible without the founding team: Geoff, Matt and Nick. While they’ve graciously described the board and investors as a dream team, the real dream team is the Meteor founders. Passionate, committed and appropriately eccentric make them the best team in the universe to be working on this project. We are honored to be partnering with Meteor in their mission to build a new platform for cloud computing applications.

We just invested $100M in GitHub. In addition to the eye-popping number, the investment breaks ground on two fronts:

  • It’s the largest investment we’ve ever made.
  • It’s the only outside investment GitHub has ever taken.

Why did we bet the farm on a series A investment? It starts with the four founders: Tom, Chris, PJ and Scott. They had a vision for a new way to develop software and created a new kind of company to pursue it. With only a handful of people in sales and marketing, the four grew the company to over 100 people, while growing revenue at nearly 300% annually—and profitably nearly the entire way.

How did they do it? They took an old technology category and turned it on its head. Source Code Management (SCM) is the second most fundamental tool for a programmer after compiler and development tools. It stores, versions and branches source code being developed by teams of programmers. At scale, these systems become highly complex and often difficult to manage. In addition, historically SCMs have been anti-social. The No. 1 conversation they generate is referred to as: “Who broke the build?” GitHub solves these two problems and dramatically expands the category by changing the old model in two important ways:

  1. Rather than forcing every development team in the world to deploy their own SCM, GitHub runs one big SCM in the cloud and the management issues vanish.
  2. GitHub organizes projects around people rather than code.

These changes may seem simple at first, but their ramifications have been stunning. Because modern programming tends to be about assembling code—in the form of libraries, open source work, etc.—as well as writing it, code tends to belong in one place where it’s easy to access. That place has become GitHub with over 3 million Git repositories.

By orienting around people rather than repositories, GitHub has become the de facto social network for programmers. If you are using another programmer’s open source libraries, are interested in what she’s doing or just a fan of her work, you can follow her on GitHub. If you need to hire great programmers, why look at resumes when you can view a candidate’s actual work on GitHub?

Beyond the growth and great products is GitHub’s incredible culture. Tom, Chris, Scott and PJ constantly push the limits on the status quo and drive new thinking in terms of management, hiring and clarity of vision. At a16z, we share this vision and I am honored to be joining the board and partnering with the company as they continue to build one of the great software success stories of our time.

When I first met Suhail Doshi, the CEO of Mixpanel, I was blown away by his passion, focus and vision for his company.  Here was a twenty-something, technical founder who articulated the perfect vision for collecting and analyzing user behavior on mobile devices.  Not only did he have vision, but Mixpanel was already collecting 4 billion actions every month and generating revenue from a wide-array of companies. (An “action” is a user-generated event that occurs when the user engages with a mobile app.) In fact, Mixpanel is on track to reach 7 billion actions by the end of May! Their solution was so differentiated and so perfect the customer base was built all by word-of-mouth.  No marketing and no sales.  Simply amazing.

It’s all about the data

Mobile applications are driving the new world of computing.  Nearly a million mobile apps have been developed in the past few years for iPhone, iPad and Android devices (TechNet).  This trend will continue—perhaps even accelerate—and the companies that make these applications are becoming acutely aware of the need to analyze user behavior and traction.  Knowing who is using an app for how long and which features are more used than others has a profound impact on how the application should be developed and deployed.  Providing a dedicated solution to this problem is exactly what Mixpanel does.  Until now, the best that mobile application developers could do was use older PC-based technology, which has been completely inappropriate for the mobile web.

It’s all about real-time simplicity

At the heart of Mixpanel’s product is a custom-built database that allows for data collection at massive scale, while at the same time offering simple analytics to users in real-time.  Mixpanel is the ultimate big data solution for analyzing everything users are doing with mobile applications.  So simple is the system to use and set up that many companies who have presented their business at Andreessen Horowitz have displayed their Mixpanel analytics at the pitch meeting to show user traction and cohort analysis.  We realized there was lightening in a bottle when nearly every company developing a web application was showing their user traction through the Mixpanel interface.  So simple, so elegant, and all without any sales effort.

It’s all about time

We at Andreessen Horowitz are ecstatic to have led Mixpanel’s Series A round of $10.25 million and I will be joining the board. Mixpanel analytics arrives exactly in time to take advantage of the massive wave of new mobile applications. We are honored to be partnering with Suhail and Mixpanel in their quest to “help the world learn about its data”.

Data underlies every Web site, gaming company, social network and enterprise computing environment. We’re talking massive quantities, too, created by users and computers, and getting bigger all the time.  In fact, we’ve run out of room to store all of the data we produce, with IDC finding that the amount of information created, captured or replicated exceeded available storage for the first time in 2007 and that the size of the digital universe this year would be tenfold the size it was just five years earlier. This is a problem that is not getting easier to solve. Solutions that enable us to more effectively store and manage data will quite possibly be one of the fundamental breakthroughs in information processing.

Enter Actifio, a fast-growing data management company whose entire purpose is to enable highly efficient storing of information. When I met the company this past fall, they presented a study that illustrated the data problem more clearly than anything else I had seen: For every piece of data created, there can be up to 12 replicas of that data that then get stored in various places. Think about that.  Twelve additional copies made for tasks such as backup, global replication, disaster recovery, data analysis, etc.  No wonder there is a data deluge problem.

Imagine the positive impact on our ability to manage information if only one copy of data was needed for all use cases. This is exactly what Actifio does. We at Andreessen Horowitz are very excited to have led their most recent round of funding.

We are equally excited to be working with one of the best storage teams in the industry. Ash Ashutosh is a technical CEO/founder who has deep domain expertise. Furthermore, the team that Ash has assembled will enable the company to scale substantially from where it is today. I am joining Actifio’s board and personally look forward to working closely with the team.

Finally, Actifio’s product is truly revolutionizing the sleepy, decades-old backup market by providing an elegant and modern solution that condenses multiple data management tasks into a single use case. Current Actifio customers describe the product as simple to use, complete, and scalable, without the expense and headache of multiple existing solutions. There’s no better validation than what customers have to say and there are a lot of happy Actifio users.

At Andreessen Horowitz, we want to make sure we are invested in the most influential software companies, regardless of stage. Actifio has the team, market opportunity, and product to be one of the most influential companies in enterprise infrastructure. Andreessen Horowitz led this recent round, joining existing investors Northbridge, Greylock and Advanced Technology Ventures. The $33.5 million Series C financing gives the company the ability to programmatically scale, ultimately bringing Actifio to a data center near you.