Ross bootstrapped Mulesoft with a paycheck and also with services. Now, the company has raised over $130 million—the last round at an $800 million valuation. Very interesting story!
Sramana Mitra: Let’s start at the very beginning of your personal journey. Where are you from? Where were your born and raised?
Ross Mason: I was born in London and raised, up until about 9, in London. Then, we moved to Wales in the UK. I was in Wales with my parents who had a hotel and eventually, hotel chains. I grew up in a business. My dad himself is entrepreneurial. I got used to dealing with customers and thinking about what people need at a very young age. From there, I went to college in Bristol where I studied Computer Science. Fairly soon after that, I got into working with banks and insurance companies. My focus was always around solving difficult problems.
Sramana Mitra: Can you put a chronological framework around this? What year were you graduating from college? What year were you starting to get into the industry?
Ross Mason: I’m at an age where I don’t even like to disclose my age anymore!
Sramana Mitra: The reason why I need to know this is to understand what was happening in the industry.
Ross Mason: That sounds plausible. I’ll go with it. I went to study in 1993 and graduated in 1997. My first job was running software for clients in an insurance company. Then I got a position in Switzerland and that’s where I started working in banking. That was probably in 1999.
Unfortunately, I missed the first tech boom, which I think would have been perfect for me. In the background, I was always trying to do my own thing and I was always enamored by the stories out of San Francisco at that time, but it didn’t feel like I had a path to get there at that time.
Sramana Mitra: How long did you stay in banking?
Ross Mason: On and off, throughout my career. Even now at MuleSoft, I still sponsor a lot of the large banking accounts as well. I’ve always stayed pretty close to it.
Sramana Mitra: As a job, how long were you in banking?
Ross Mason: Probably about four to five years.
Sramana Mitra: This brings us to 2003?
Ross Mason: Yes.
Sramana Mitra: You were on the technology side of banking, right?
Ross Mason: Absolutely, I was a technologist. I was working with the heavyweight back-end systems and trying to move the information around their systems.
Sramana Mitra: What happens after you’re done with banking? Is that when you started MuleSoft?
Ross Mason: I started in 2003. I worked at a pretty good and fairly well-known project in the banking sector for one of the investment banks in London. This is really where I cut my teeth on what we do today. It was a pretty astounding project and by today’s standard, doesn’t even sound that big. Back then, we were trying to connect systems. It took about €30 million and 18 months.
What really struck me were two things. One was how woefully complicated it was to deal in that environment. I couldn’t understand how it could have been the way it was. I was in the middle of it, but I just couldn’t fully grasp it. The second thing was I realized a lot of the architectures and concepts we were dealing with were very sound. I really liked the notion of how we solved some of the problems. I did not like execution and the type of software available to realize those great ideas. That’s what spurred me into creating MuleSoft—to create a platform technology that enabled these new type of architectures but in a way that was more friendly to developers.
Sramana Mitra: How exactly did you architect MuleSoft? How did you think about what to put out there? You obviously knew the problem that you’re trying to solve from your experience. How did you architect the solution?
Ross Mason: I looked at app server. App servers were getting really big then. I went to servers because open source was driving new frameworks. In fact, the first incarnation of MuleSoft was based on app servers and it was terrible. It ran like a dog. It was very cumbersome and it didn’t hit any of my checkboxes of what I was trying to do.
Then I stumbled across a Master’s graduate paper from a guy called Matt Welsh. He was touting this type of architecture called staged event-driven architecture. It was to build high-scale web service at that time. I realized the architecture could be applied to any kind of distributed system. In the very early days when I spoke about MuleSoft, the architecture was a key principle of it. The system was architected to be decoupled and scaled linearly across commodity hardware. I started around that idea of just creating a brokerage system that would allow you to send any type of event in and out using this staged event-driven architecture approach.
Sramana Mitra: You were going to use this as glue to do the kind of integrations that you struggled with at the bank.
Ross Mason: Exactly. The idea was we’d have the underlying architecture to get the performance and throughput. On top of that, you’d layer a set of well-defined concepts of how you would stitch applications together. They were not unheard of concepts but just making them very well understood inside the platform.