JOHANNESBURG — When thinking about top South Africans in San Francisco’s Silicon Valley, the first names that may come to mind are the likes of Elon Musk and Vinny Lingham. However, there’s another South African connection who has literally worked at the top levels of major tech giants in the US, building out their engineering systems and helping to create the cloud platforms that the modern business world benefits from today. Chris Pinkham, who was born in Singapore but moved to South Africa when he was 9 years old, was behind South Africa’s first ever internet service provider Internet Africa. He also helped pioneer the first-ever internet connections in South Africa while studying at UCT. He went on to lead engineering teams at Amazon, working in Seattle – and even briefly in Cape Town – eventually becoming Amazon’s VP of engineering. He would also go on to become VP of Engineering at Twitter (a role he left only earlier this year), helping to solve some of that company’s biggest technical challenges while also previously selling another business of his to Oracle. Pinkham, who now resides in San Francisco and who recently turned 50, speaks to BizNews about his star-studded technology career and how the rapid changes in South Africa in the 1980s and 1990s helped shape his worldview. Today, Pinkham is getting involved in South Africa’s technology space once again, this time by joining the board of a top, up-and-coming Cape Town based FinTech disruptor called JUMO, which is unlocking the potential of the cloud to deliver inclusive finance services particularly in developing parts of the globe. – Gareth van Zyl
I’m speaking to Chris Pinkham via Skype from the Bay Area in San Francisco. Chris, you’re a former UCT student who has literally worked your way up to the top echelons of the tech world in the US. You’ve served in really high roles at Amazon and Twitter, where you were VP of Engineering for both companies. For our listeners and readers out there can you tell us about how your career started?
Sure, I was very fortunate to land my first job at the University of Cape Town in the late 80’s when connecting university campuses was a key need for our researchers. This is pre-internet days and based on my interest in computer networking I got stuck into those projects and with collaboration and help with the colleagues at other universities. We stitched together the first internet in South Africa, the first network connections outside of South Africa about 24 or 25 years ago. When I left UCT after a stint in the US I came back and with a couple of those friends I’ve worked with before we decided this internet thing had legs and was important and needed, and we would set about building what was the first internet service for the general public in South Africa, which launched in 1993. It was a product born out of need and out of interest and it just ran away with itself. It just had enormous momentum. It sounds kind of silly and trite saying this 25 years later, but it was a very uncertain period. It’s just astonishing to see how things that are so necessary can grow so rapidly.
What’s the whole back story behind Internet Africa and especially the story behind you selling it? It’s fascinating that you started essentially the first internet service provider in this country. Obviously, the market in South Africa is much smaller than elsewhere, but it was a very interesting time, as you say. It was the early 90’s and there was a lot of change in South Africa.
Yes, it was a very interesting time at the time there was a telco monopoly in the hands of Telkom. We weren’t certain what we were doing was legal. There was talk at one point of letters being issued to shut down our international connection given the political dynamic at the time. You must remember this was pre-1994 and before full democracy in South Africa, so there was a lot of change a lot of flux but the original back story was really a combination of the points of view of myself and a couple of others, who had worked on the academic networking setup in South Africa, and just realising there was this enormous need and enormous opportunity to provide relatively, affordable, generally free and open internet access to the general public. At that time, there were very slow modem connections. We were in a fortunate position. We had the experience and the skills and the insights, and some of the opinions on how this gets done and it gets done rapidly so that people could have access.
Part of it was informed by – I spent a year working in the United States where the internet was being commercialised. It was still in the early days, pre-world wide web days. And coming back it just seemed like a pretty clear opportunity. So we seized that in a very, kind of bootstrap mode using very low-tech networking gear for the time, open sourced software, we scavenged equipment and favours from various technology partners and we scrambled to get funding together, in order to be able to afford internet connectivity, which was the grand sum of 64 kilobits per second. Which is about a thousandth of what a reasonable internet connection is considered to be these days for one house, so that was the whole country.
Yes, it was a fun time, it was a crazy time. We did the best we could with what we had but it all worked out in the end. We took one step after another. We had no idea what we were doing, from a business point of view or business planning funding point of view. It was entirely customer driven and we were just fortunate we developed a service that people needed and flocked too without a lot of pushing.
Tell us about what happened to your career after you left Internet Africa, after you guys sold it? What was the trajectory of your career from there?
I took a bit of time off. My wife and I went sailing for a year and along the way, we had our first child and after all of that I was ready for the next challenge and then Amazon came calling. I had a friend who I had worked with at Internet Africa, who had moved to Seattle and he put Amazon in touch with me and they hired me to come and run network engineering in Seattle at the end of 2000. That was a pretty interesting time to join a company like Amazon, it was towards the end of the DotCom bubble. Amazon had reached enormous scale but it suffered a lot of growing pains and it was on a path to getting to profitability and there was a lot of fixing of technology issues and people issues and funding issues, for the services we were managing. So that was the next step and I spent 7 years in Seattle, came to run all of the global infrastructure before deciding our hearts and family were back in Cape Town and we wanted to spend a little bit more time there, as our kids were growing up in South Africa. Christine and I decided to return to Cape Town, which opened up a whole new chapter.
Were you born in Cape Town, Chris?
I was not, I was born in Singapore so I’m a child of a military family, which moved around the world until we landed up in South Africa when I was 9.
Do you then perhaps have South African citizenship?
I have a very mixed background but I do have South African citizenship, an interesting back story on that was in the 1980s the South African Government decided to force naturalise or children of immigrants so that they become eligible for National Service, so we couldn’t use the loophole of not being South African to avoid the call-up. I’m very happy to be a South African but the way it came about was just interesting in a kind of macro-political context.
Just getting back to Amazon, you essentially are accredited with having built Amazon EC2, which is described as an elastic cloud computing service. Obviously, cloud computing and Amazon have become synonymous with each other in many respects. Amazon has come to lead the world in terms of how to do cloud computing. Can you tell us a little bit more about the story around that and how you feel about where cloud computing is going now, the direction that it’s headed towards?
Sure, how it came about is a multipart story. It’s partly a story of Amazon Web Services originally focused on how to bring e-commerce services to its developer community, which was fairly substantial, before it was involved in the cloud computing business. It combined with other internal learnings and architectural decisions for Amazon data centres and Amazon developers. Thirdly, the catalyst for this was my decision to move back to Cape Town and the Amazon CEO asked that I not leave the company and we talked about what it was that I might work on and if I would look into this compute service that was a goal of the company at some point in the future. It was un-staffed and undefined, so with some of the folks that I worked with in Seattle, we put together a proposal, which described in very rudimentary terms the first compute service, the elastic compute cloud. Then we sought approval to return to South Africa and hire from the Cape Town community (South African community), a number of engineers to go and work on this. We did that in late 2004 or early 2005. We had our first engineers on board in the Constantia area in Cape Town.
This was the genesis of the Amazon presence in South Africa and this team worked on developing EC2. It was launched in 2006, so I think from an importance point of view the problem we were trying to solve was how do you free the developer of a service – and Jumo is an interesting example of this – how do you free the developer of a service from having to worry about all of the nuts and bolts and the paying and planning and scaling and operating large scale infrastructure? How do you allow them to spend more of their time and resources on the thing that they set out to do rather than the common infrastructure that is for them, relatively and differentiated? How do you do that in a way that is economical and allows them to respond to demand and match their demand of their service against the demands of the infrastructure more closely?
So, I think EC2 was a pioneering offering in that respect. I don’t think there was anything quite like it. There are now several competitors to Amazon, but that first mover advantage has certainly put it in a leading position. Where it’s heading – there’s still a lot of work to do in terms of being able to translate a developers’ creative idea into a service that’s available to their customers almost immediately. In the way I’ve tried to describe it, in groups that I’ve worked with, is you really want a developer to be able to express a thought and have the benefit of that thought available in as little time as possible and hopefully no longer than the time it takes to test and deploy, and that should certainly be sub-hours, preferably close to minutes.
We’re still a fair amount away from that where we’re still operating at a machine and network level infrastructure, rather than a scalable set of developer tools that operate more at the kind of the speed of thought of the developer. Yes, I think we’re getting there. I think the layers are being built slowly, which allows that to happen but there’s still too much gunk in the way.
You were also a VP of engineering at Twitter. What was that experience like?
It was an amazing experience. I’ve been a big fan of Twitter for about 8 or 9 years or so, and more as a lurker, or more as a reader. It’s just an incredible way of having a real-time view on what’s happening in the world in a way that is not represented through other media, so it’s an incredibly important company on a global stage, from a social point of view. I was super pleased when I got contacted and asked if I would come and join and come and work on some of these infrastructural challenges that the company was facing. Like most successful companies there’s a lot of scrambling and a lot of pain and thrashing but it’s not necessarily visible to the users of the services.
Sometimes these things looks like it’s easy. Sometimes it looks like everything is just happening so smoothly but behind the scenes there’s usually 100s and 1000s of people who are scrambling every day to keep things working, to scale well, to secure the system. Again, it’s all about serving the customer. It’s all about trying to figure out how to increase engagement and activity on the platform by providing what the customer needs, (even though they don’t need it in the moment) so it was a great experience. It was a fun time working at one of the most progressive but still very young new media company, so yes, I had a very good time.
Why did you leave Twitter? There were reports that came out at the time saying that you left quite quietly. What was your rationale behind leaving that social network?
I always do everything quietly, you’ll see that I just try and be a little bit of a mystery man. The rationale, I think I just reached a point in time where focusing all of what I do on one company just doesn’t sufficiently interest me, personally. I think I’m fortunate in that I can take a step back and spend some time on maybe getting involved in multiple activities that are personally interesting, as well as professionally useful, and I find the model of being in a company, a corporation full-time, devoted just to that sole purpose – it doesn’t really exercise my brain and I don’t think it’s the best use of all of my time. I’m at a point – you know I turned 50 this year – and I’m just at a point where I want to spend time doing lots of things that I find fascinating. Whether those things are Silicon Valley technology related or have nothing to do with technology and having the time and freedom to choose what I do is something that I’m valuing right now. That was really the primary motivation.
The secondary motivation is I had a hell of a commute and that just didn’t add to the pleasure of waking up every day and getting stuck in. So, yes, I’m excited to have a whiteboard in front of me that isn’t cluttered and I can get to work on whatever I want to in more of a multiplex kind of way, so I’m very excited about that.
Your career, as I said earlier, is like a snapshot of all the big players in Silicon Valley. You also founded a company called Nimbula, which was snapped up by Oracle. Can you tell us a little bit more about that as well, and what happened there?
Sure, so my partner at Nimbula, Willem van Biljon and I, we’ve known each other since the Internet Africa days. He was very helpful with some of the technology components that we were securing for that service, so he and I stayed in touch. We worked together on EC2 when I returned to Cape Town and I thought he’d be a great partner to help build the team and help define and build and take to market the service. We worked on EC2 together in Cape Town. After we both left Amazon we realised… We wanted to do something together, we felt like we had at least one start-up in our future and we’d like to work together on that. We realised that the push into public cloud service meant that customers of public cloud services were going to get a significant load up and advantage on developers and IT managers inside enterprises, who were going to get left behind, in terms of the quality and the responsiveness of the infrastructure tool being available to them. So, we built Nimbula to adjust part of that. It was a systems level software play design to allow IT operators to automate their infrastructure and to get their developers a much more responsive experience. Similar to what they might obtain in a public cloud service.
We built that. We raised money from Silicon Valley, venture capital partners. I moved back to California to run that company and we worked on this for a little while. We found a few interesting things. One is that the market for this kind of thing has developed extremely slowly. There’s tension, put it that way, within the enterprise infrastructure community on what the best approach is. We found that we learnt some lessons in terms of how many changes you can make in an enterprise infrastructure in order to achieve a successful sale. We didn’t have the business success that we hoped for. We didn’t have incredible adoption, we had some very valuable customers. Yandex the big search engine in Russia, Orbitz – the travel website in the United States, and several others.
Ultimately we decided that that’s fit for this company was to be inside one of the big enterprise technology companies that had a much longer runway and the ability to fold this into existing enterprise tooling. So we decided to pivot and seek a tactical exit on that basis and we were very pleased with the outcome to sell the company to Oracle, who had a great need to modernise their infrastructure management capabilities. They were interested in having a combined public/private cloud service which suited our software very well. As of today the software that we built at Nimbula and its successors being built by engineers at Oracle, by powers Oracle’s private cloud as well as their public cloud infrastructure services.
It wasn’t a massively successful business to be candid. We learnt a lot of lessons along the way and had some fun and pulled out lots of hair. Ultimately though, the original vision was a good one and it’s working very well. It just happens to be working very well in an Oracle context and has really enabled Oracle to push into a newfound relevance in the public and private cloud infrastructure as a service world, so it was a great exercise. It was one of the things you have to accept when you start a business is that the future is extremely unpredictable and the only result from any plans you make is that the result will not be anything like the plan you predicted, so you’ve got to be able to adapt, pivot, and respond to the market conditions that you see in front of us.
It was a great learning experience. I think the important thing was that the smart people who are working on this are still making a huge impact on the development of this technology and these services today, so we had a lot of fun.
Chris, fast forwarding to 2017, it almost looks like you’ve come full circle in a way. You’re involved in another Cape Town venture called JUMO. You are now on their board, correct?
What attracted you to JUMO in particular?
I was very attracted to the team of people that I met. Starting with Andrew Watkins-Ball (JUMO CEO and founder) and then subsequently his team operating out of Cape Town. I was particularly attracted to the problem they’re trying to solve, which is how do you bring affordable financial services to people who generally have been ignored by the financial community. The way they are going about that is very smart and they seem to have hit a certain inflexion point, in terms of product-market fit, which means they’re having enormous impacts. So, a combination of all of those things, it fits my own personal business plan of spending time in multiple projects and I just couldn’t be more thrilled to be working with such a great bunch of people who are working on a problem that just is going to have enormous, global impact.
So JUMO is really about inclusive finance? Getting more people, specifically in developing parts of the work to be able access finance, is that it?
Exactly. I think something like 80% of JUMO’s customers are small entrepreneurs who needs small amounts of capital in order to build up their business and this could be as simple as equipment needed for a hairdressing studio, or to buy inventory for the days market or whatever it is. The reason these communities have been underserved is just incredibly expensive to bring this tool into these places and to develop a predictive credit risk assessment. And JUMO’s insight is how to turn these customers’ mobile behaviour into credit scores that represent the predictive value of their ability to afford this credit and to do that in a way that has the lowest unit cost in the marketplace. Being able to do that they are now able to partner with mobile operators and banks and other partners to provide a good view on what kind of facilities these financial partners should be able to offer to these customers.
They’ve been very successful with this. They’re expanding but they’ve grown very healthily in Sub-Saharan Africa, and now expanding into Asia. The market size available in this kind of marketplace is in many 100s of millions of potential customers, so it’s a very exciting time for them and as they go through this expansion they need help trying to figure out how to execute well, how to scale their business and having an old greybeard like me, who has some scars and has maybe seen a few of these problems. But also just has a general perspective, it just seems like a pretty good match.
What kind of help are you going to give them because I presume that they need a lot of insight, in terms of their cloud background especially? Cloud will be key to them.
They’re a very cloud-centric company. They rely very heavily on the public cloud services available to them. Philosophically, it works very well for an emerging market company who is situated very remote from the technology centres of the world. To have access to these world-class infrastructure facilities. They make really good use of those, so a little bit of help on how to do that but they’re experts themselves. They’ve done their homework, they’ve spent years now building and really have a great understanding of how to make best use of these kinds of tools.
My contribution is a little bit of take a step back and strategically think a little bit more long-term. Maybe introduce some things that I’ve seen, maybe techniques, maybe people, or maybe implementations. Challenge people not get too locked into their best assumptions. But these are very good engineers. It’s astonishing the quality of the team that’s been assembled. I generally assist more in terms of high-level strategic thinking. A lit bit of review, a little bit of roadmap and product assessment once in a while, just to bounce ideas around. Make some introductions. I’ve met some smart people at places like Google and Amazon, and Twitter and occasionally making introductions to allow people to think a little bit beyond the horizon.
One thing I’ve noticed, having operated both in Silicon Valley and in Cape Town – their points of view are roughly the same but sometimes it’s very hard to kind of get out of a very localised way of thinking. In Silicon Valley, we refer it to as the Silicon Valley bubble where we assume that all of the problems in the world are the problems that Silicon Valley is having. Like how do we get our lunch delivered cheaply and freely? But being a developer in a city like Cape Town or even more remote part of the world it is sometimes very hard to connect in real time with what folks in other parts of the world are thinking and it turns out. No matter what problem you’re working on the odds are somebody else who’s also very smart and who’s been hit by that problem before and maybe able to offer some insight. So part of what I do is try and step back and try and connect people. Try and connect ideas and just try and be a bit of a new sight. It can be both in a product technology domain as well as the life of a start-up and being an entrepreneur.
It seems like Cape Town has had quite an interesting start-up culture of late. What do you think is the standard of tech start-ups in Cape Town, currently especially seeing as you got involved with the likes of JUMO? Do you think you could find another JUMO potentially in Cape Town?
I think it’s possible. Certainly, the standard of individuals and the standard of talent is as high as anywhere in the world. I’ve been looking at this since 2004 when we were thinking of bringing Amazon back to Cape Town, one of the challenges we had and one of the key strategic discussion points that we considered was what is the level of talent and how could that scale? There’s no doubt that the level of talent is extremely high, as was demonstrated with the Amazon project. The question is more what is the scale of the available talent? It’s improved a great deal but it’s still a very small talent pool on a global scale, so just the numbers are hard. The successful companies often happen because a hundred attempts were made at trying to build a successful company, and to build a hundred successful companies, you need thousands of talented people, so it’s certainly possible. There’s certainly no great reason fundamentally why there can’t be many JUMOs and takealot.coms and other successful companies.
For there to be many of them the available talent pool needs to be drastically expanded and that’s no small challenge. That goes back to educational issues. That goes back to science, technology, engineering post-graduate programs. Over the last 12 years, it certainly seems like there’s been enormous progress. Many more people joining programs and Cape Town being this hub, which is attracting people, which is having a virtuous cycle effect because of the availability of interesting projects more people are tooling up and getting involved in those. I definitely see the possibility for more JUMOs to come to the market, in order for them to succeed they’re going to need resources like funding, which is still relatively immature, it’s certainly improved but I would say it’s still in the very early stages.
Bearing in mind that Silicon Valley has been for probably 50 years in the making, from a venture capital start-up basis. The model has not been replicated successfully anywhere else. There are little pockets of it in London, Boston, and New York City but very few Dollars, relative to what’s been allocated in Silicon Valley are being allocated to these centres. There are significant uphill challenges but there’s no reason why an entrepreneur, a creative individual couldn’t start whatever they wanted to in South Africa. The infrastructure barriers have largely been withdrawn.
Apart from JUMO, what else are you involved with nowadays, especially there in San Francisco?
I’m having multiple conversations with an early stage, pre-funding, pre-formation technology pre-company that is trying to figure out if they’re even going to do what they’re were going to do. I’m also having conversations with one large technology company. I’m trying to see if there’s any way I can help them figure out how to be more relevant in the public cloud space, so other than that I’m engaged in several other interesting pursuits. I’m spending a lot more time on my bicycle, which I used to spend a lot of time on in Cape Town. When I got into the Silicon Valley lifestyle I neglected, so I’m doing a great combination of professional and private pursuits.
And could we see you back in South Africa, settling here again or are you accustomed to the Valley now?
I’m pretty settled in California, and I have 2 kids in the local high school, so we are not going anywhere anytime soon. But I’m very happy to report that I will be back in Cape Town several times a year on business, and I thoroughly look forward to doing that.
Chris, it’s been an absolute pleasure talking to you today, very insightful. Thanks for taking the time to speak to me and to our audience.
It’s been a great pleasure. Thank you very much.