“Hacking and assembling and and that kind of thing? It’s kind of what I do in my sleep.”
Every Monday morning we fire five questions at a C-suite tech industry interviewee. Today we’re pleased to be joined by MuleSoft CTO Uri Sarid.
Uri: What’s the Biggest Challenge for your Clients?
The sheer number and variety of systems that they need to make work together. The diversity of the people trying to put them together; the diversity of the systems; the diversity of the needs and the almost despair – that by the time they have finished hooking those systems together, ‘we’re going to have another load of projects’. Our customers’s biggest challenge is ‘how do we ever get out of this mess and actually start adding business value?’
[From our perspective] what we needed to do, is to find some way that doesn’t force all the systems to be alike in order for them to work together; to have a product on our side that can work with essentially any system on the other side – and do it in such a way that the skills that the customer needs in order to make things work are the same, regardless of what system they’re hooking up to.
The main idea was to put productised API – really well-defined, clean APIs – in front of every system out there; make it really, really easy to connect to them and create an API in front of them. And then of course, once you got that productised API, to make it very simple to connect the APIs together. That’s turned very much into a lather, rinse, repeat, kind of problem; which is exactly what the clients were looking for.
Technology that Excites You Most?
I’m super excited, as is everybody else, around the promise of Artificial Intelligence and Machine Learning, and how can you apply it.
I’ve also done a lot of work on IoT – connecting the physical and the virtual world, and that’s super exciting as well.
New mapping technologies too: the ability to use drones to do both mapping, and delivery, and repair and so on.
Mapping is interesting because it is a precursor of IoT. If we want to – for example – turn the city into a smart city that actually works better for the people inside of it, then knowing where all the physical, static assets are: that’s an ideal capability for turning the city into a smarter one. To do that, location technologies and mapping technologies are absolutely critical.
The real question that I’ve been asking myself is, how do we make it easy to take advantage of all these technologies? That’s the big question still.
Putting aside the obvious greatest success really, which is home and family* and leading a fulfilling life, I think it is [helping create] an integration technology.
This notion of connecting everything together; the issue was seen as effectively intractable. How can you possibly do something that makes everything work together when you don’t even know what those systems are? I think we’ve created basically a new category in trying to solve that that problem. That was not trivial.
The solution involved something akin to the computer network, but really at the application level, called an application network. There’s an underlying graph that actually captures all the information; the wiring diagram if you like of how it all works together. And it actually works! This isn’t a theory: Airbus and Amtrak and, you know, various airlines are actually using it for their mission critical systems.
*I have a wife, I have almost 17-year-old son, who’s going to go off to college in a little over a year. It’s going to be heartbreaking. But you know it’s a good thing…)
I think the failures are in some deep sense as important, if not more so, than the successes. If I look back in my career, and I look at the at the business ventures and the companies and the things I’ve tried to do that have not worked out the way that I wanted them to, there’s a long string of them!
A lot of things have to align in order for you to be successful. You have to have the right funding, you have the right market timing; there has to be a technology fit between what’s working in the market and what is actually needed.
For example I had a startup that tried to solve a really important problem around task management, but it turned out that it was a peer-to-peer architecture, and the market was looking for web architecture.
A lot of things have to align, and when they do align you can get a spectacular success. Even when they don’t align, perhaps that technology was right but the market for it was not. For example the work that we did here at MuleSoft around RAML and API specs really dated from a couple of companies back where we saw the same need. We saw that it actually worked but it wasn’t the right time to actually productise that; I took that learning and applying it here at Mulesoft.
In Another Life I’d Be…
I asked myself that about 20 years ago when I was in physics: what would I be doing if I wasn’t in physics? The answer was computers.
So now I’m doing pretty much my hobby as a full-time thing. (If you ask my wife, she’ll tell you that I spend the weekend hooking up connected cars and integrating our home system…)
I also like reading and kayaking and a few other things, but hacking and assembling and and that kind of thing? It’s kind of what I do in my sleep.