Dial Up Motive

Hello, hello, everyone. Welcome to episode 34 of Dial Up Motive. I am your host, Dan Hollinger, dialing in from the Munich coast.

If you've never caught this show before, this is where I interview Cloudflare employees about their earliest experiences with technology and the Internet, and really how learning how it helped drive them throughout their career and where they ended up at Cloudflare today.

With me, I am joined by Andy Kennedy, a senior manager for solutions engineering here at Cloudflare reporting out of London.

Andy, thank you for joining me today.

Hey Dan, I'm delighted to be on my first Cloudflare TV segment. So yeah, I'm delighted to be here.

Awesome, glad to have you. And how long have you been with Cloudflare?

This is your first Cloudflare TV segment. Yeah, just under four months. So yeah, it's been some crazy time, but I enjoyed every minute of it.

So yeah, delighted to be part of today's session.

Awesome. Well, we're glad to have you. And could you give us some insight into what it takes to be a manager of the solutions engineering team or what the solutions engineering team does here at Cloudflare?

Sure, happy to.

So the solutions engineering team is focused really in helping our clients use our technology to accomplish a particular business outcome, to solve a problem that they face.

And if we think about the needs and the challenges that organizations of all sizes face today, whether that be around application performance optimization, whether it be around security or connectivity or remote or hybrid working, the team that I get the pleasure to support are technical experts who can really lean in to understand, okay, how do we help you meet that particular outcome or solve that problem in a way that's efficient, that makes sure that we're open and we're able to support them in whatever the circumstances that they face.

So yeah, so it's a great team. I'm delighted to be part of them.

And yeah, I'm very proud of the work that we do. Awesome. Thank you for that context.

And so to dive right into the show and the heart of the show, as you can see behind me, I have not my desktop, but really the spiritual successor, spiritual representation of my desktop in the era of Packard Bell and CD -ROMs and at least the little floppy disks, if not the big floppy disks and the early Internet.

And what I hope to learn today is what your first or earliest experience was with a piece of tech that really shaped your life or shaped your career moving forward.

So happy to kind of jump into that story. Sure. So this goes back probably to the late 80s and all of my friends, they had Spectrum or Sinclair Spectrum computers, or they had Commodores or whatever it might be.

My brother actually happened to be a computer technician at the time, and he happened to bring home a very early PC.

And it was a great opportunity for me to experiment and learn whilst he was kind of out working on what this new thing was.

And the first impression was it was loud.

I mean, this thing had fans and must have consumed power and put out like nobody's business, but it was a DOS-based PC.

So it had a green screen, so still kind of black and white effectively, and ran the word processor Multimate.

So this is the pre kind of Word era. And I remember distinctly the fact that it's something called Norton File Commander and a whole bunch of other kind of utilities that I really didn't understand or have a clue about.

But the one thing it did have was a copy of Space Quest 2, which was this kind of role-playing game.

And I tell you, it caught the imagination in terms of just thinking as a child, the fact that you read books and everything else, but now you were experiencing it, you were interacting with it, you were getting to a point of discovery.

And really from that journey, from that day, it started. And that was kind of my first experience with computers and the idea that you could control the thing that you were interacting with to do stuff that was basically the limits of your imagination.

Now, was this one of the early text-based RPGs where you were given a prompt and then writing an action and then it would react to that and carry you through the story?

Or how did the game play progress? Yeah, well, it was pretty early graphics.

I mean, I think people would laugh at it now, but it was text or block-based graphics.

And of course, there was no joystick, there was no mouse.

So you were kind of navigating using the cursor keys. But yeah, even with the most basic of what we might say are graphics today, it was still something that you could render it in your mind and you could see what the authors of the game were trying to portray.

And of course, the sound was just a beep.

All they had was there was no sound card, there was nothing. It was just beeps and kind of noises that came from the machine.

So if you could hear over the fans, yeah, the beeps were there.

And that was your world of immersion then. So it was a great experience and I'll look back on it with fondness.

Yeah, it's always fascinating looking back on some of the ancient games that as a kid, you remember them being super immersive and the graphics were hyper-realistic.

And you applied so much of your imagination to the world that was being displayed on the screen.

And now when you look back and you're like, oh, that was a pixelated pile of like nonsense, I can barely understand where my character is, let alone the ball or the spaceship.

And it's fascinating just taking that trip back and be like, oh, because I think I grew up with the Atari and some of the earliest kind of graphics.

I'm like, oh yeah, those were impressive at the time, but fascinating just how much you can immerse yourself as a kid into that character, into that role and scene.

And I think the thing that maybe we miss, and maybe come back to this later in our conversation, was the focus on the user.

I remember before, actually, that was my first computer.

I guess my first experience of games was, I remember a family friend having, do you know the table tennis game where you've got the two, is it pong?

You have the two things. Ping pong. Yeah. And then pong, the digital version.

And I remember playing that, but again, you could sit in that for hours and your excitement and enjoyment you got from it, because I think it really put the user first and you were like, this is, I'm playing the game here, I'm in control of this.

And it was an amazing experience. And I think it sounds a bit silly now, because if you were to put that in front of people, they would kind of laugh at it.

But again, caught the imagination, got me excited. And I think I'm grateful for the fact that that was something I had the chance to grow up on, because it's really a formative part of where I am today.

And how did that excitement for that early game or that early experience with the computer kind of continue to evolve as you grew up or continue to pursue new interests?

Well, the funny thing is that I've talked about games, but actually what ended up happening was over the years, I was fortunate enough, maybe we got a little bit more of a modern computer again, family member was kind of learning on these things.

So I had the chance to borrow theirs. And we eventually had a Windows system.

And for me, it no longer became about playing the game, but more actually about getting the thing to actually work.

And my curiosity was more around dealing with the problems back in those days of just memory management.

You had to put into certain memory modes to get certain games to run.

And it began to become more of a curiosity about what was happening behind the scenes.

How were these things running?

What was our batch file? What was contained within it? And what happens if you begin to change some of the parameters?

Would something happen differently?

More often, it would end up in tears and something would break, but then you would kind of fix it and figure it out.

And of course, there was no Internet really accessible at that time.

So you were just trying to be curious about the mechanics of this thing that you have in front of you.

And again, there was nobody really telling you no, or it doesn't work that way.

It was just more of a case of tinkering around the system.

And then you would go into discovery. You would go into a folder, you'd have no idea what these things were.

You'd just open them up or you'd run the command and just see what it did.

And it again, piqued that curiosity.

So as I said, it became less about gameplay. And even today, I don't really play games, but it was more the fact that just kind of getting these things to run was the achievement in its own right.

And yeah, that's one thing that's been fascinating about some of these early conversations is, as computers were becoming more popular, there was still that educational gap where you had to go to the library to pick up a book, to learn about the Internet, or to learn a programming language.

There wasn't this just fast access to the public Internet and knowledge the way it came about later.

And so I've had conversations where people had to go pick up the book on HTML or C++ or even earlier languages than that, early C to actually start to interact and play with their computer.

And as you were messing around with these folders and really exploring, what was one of the takeaways from that other than a few broken applications or reinstalled operating systems?

Well, it was, that was kind of, look, I mean, I think, just remember what you did before, so you can go back and correct it, because it was, it was a case of, I guess, an early version of version control, because you would make a copy of the file, if you messed up, I just reloaded the file back in and things would work again.

So you kind of got taught the basics of systems administration by accident.

And, you know, I like your kind of comment about getting the books.

I mean, for me, actually, there was a very popular magazine here in the United Kingdom called Personal Computer World.

Unfortunately, it's long, long gone, but it used to be 90% adverts, maybe.

And it came, it was about this thick, huge thing.

And it always came with the disks taped to the front. And one of them would be probably for some form of dial-up software that you would need to connect to like an ISP.

But there would always be kind of shareware or freeware type utilities at the front.

And yeah, it was always kind of learning experiment with them.

And, you know, it keeps me kind of, you know, fond memories of installing like Doom for the first time and everything else that came on those magazines.

And the magazines were brilliant because there would always be a little snippets in them with basic programs.

So I remember GW Basic and QBasic and these other kind of utilities, and then more modern kind of languages, learning a little bit about basic programming.

And again, it wasn't done in an academic way.

It was kind of, you were basically going through the magazine, you'd come to a section, you'd go and try something, see if it worked, and then you'd go on to the next thing.

And of course, back in those days, I don't think there was as much distinction between the corporate world and, you know, the personal interest, because it used to blend in kind of developments around things like Windows NT or other forms of operating systems like OS2 work.

And all these other things would be mixed in with consumer articles about graphics processing on desktop publishing, which was all the rage back then, you know, these kind of ideas.

So yeah, it was a fascinating time. And one I look back definitely with fondness as well.

Yeah, and just fascinating to think how far we've come.

So now anyone can share code, download code via GitHub and get going. But before you literally had to order your code in the mail, type it up or add it to your machine in order to, you know, play a game or whatever the software was sharing.

And it's fascinating just seeing the difference.

So, you know, as you went from an early game experience to starting to tinker with these new machines, is this around the phase that your household started to get online or you had Internet at school?

What did that look like? Well, actually, there's two steps. The first, when I went to university, that was the first time I'd ever had Internet access.

And I did a degree in electronic and electrical engineering, but I did it with a specialism in audio.

So I was looking at things like how digital audio converters were, digital analog converters rather, were created and amplifiers and how CDs and CD players worked, et cetera, and digital communications.

And we actually in the laboratory that we had, we had Next workstations.

So Next Step, I guess, was the operating system and Next was the company Steve Jobs founded actually when he left Apple.

And it was also the same systems upon which I believe Tim Berners -Lee actually created what we now recognize as the World Wide Web.

And it was a system that was so far ahead of its time.

We had access to email, access to the Internet.

And back then you could actually open the email system on a Next workstation and you could actually do things like integrated voicemail.

So you could click on a little icon when you were creating your email and you could actually record your voice.

Because again, the Next platform had great digital signal processing, great analog to digital conversion, et cetera.

And it would actually record the audio.

So as part of your email, you could actually have a voicemail that would go with it.

And you think about that, that's the kind of mid -90s, early 90s.

So way ahead of its time, even today. I mean, you still don't have those capabilities really in many email platforms.

So yeah, I mean, for me, it was that experience of being able to be connected, suddenly being able to do discovery.

And for me, my passion was music and being able to discover new music and understand a little bit of what was happening in a way that was never possible before.

It really did spark the imagination. But then I left and maybe my first experience of connectivity and the Internet rather than the World Wide Web, but the Internet more specifically, was when I happened to be an apprentice at IBM.

So I was working in their server development laboratories. And one day we were asked to basically look at a program to do terminal services or thin client computing as it was called at the time.

And it was the precursor to something that most people recognize today as Citrix.

But back then IBM, there was still a debate between token ring and ethernet.

So the debate was still raging about which was the better technology.

But we had to connect effectively a server to these thin clients.

And for the first time I had that opportunity in that light bulb moment when you could connect two devices together and you could make them talk.

And in that moment, it started a passion for networking and trying to understand a little bit about how these machines talk to each other and communicate.

That was the thing that led me on to, I guess, a career in networking and security and ultimately to where I am today at Cloudflare.

Because that moment of being able to understand things like IP addresses, being able to understand a little bit about connectivity models, what's the difference between a unicast and broadcast model, all of those things began to drop.

And ultimately, at that point, that set me on my course or direction.

I went back to university and graduated really based on that moment of that networking clarity.

That's fascinating.

And for our younger viewers, would you mind giving a brief description of TokenRing or how it compared and contrasted against the protocol that ended up winning out?

Yeah. Well, if you think about it, I'm trying to think how to do the analysis.

But if we think about TokenRing, the fact that with Ethernet, there's a broadcast medium, so you'd have collisions and everything else, whereas TokenRing, you basically, I'm trying to remember, there's a protocol data unit you put onto the ring and effectively it'd be carried.

And it'd run at 16 megabits per second.

It wouldn't have some of the same issues with the conventional broadcast medium, if my memory serves me right.

But ultimately, Ethernet won out. It was cheaper, more cost effective, more prevalent, kind of like the Betamax versus VHS days.

And again, that may not necessarily mean much to people. But basically, in video formats, there was two.

Betamax might have been the more technically better one, but it was VHS that won out because of its availability.

And ultimately, I guess, there's a lesson there to be learned in terms of accessibility of technology.

It's more accessible, it's probably going to be more successful.

Ultimately, Ethernet won out, and that became the feature. I'll probably admit it.

I was going to say, I didn't mean to put you on the spot, but the last time I heard TokenRing, I'm pretty sure it was from my own networking book in college and just like a paragraph of like, yeah, there were these two competing technologies and this one lost.

And so I'm like, OK, that's what I'm going to commit to my memory, and then we'll move on.

So after actually getting your teeth sunk into the Internet in college and really starting to understand networking or learn about the power that comes with starting to connect these computers together, where did that lead into either your early career or your first job?

Yeah, so I started out actually at Goldman Sachs, the investment bank.

Oh, you immediately just like, was it, were a trader and just went to finance.

So is that it? I wish that were true.

I wish that were true. No, I was part of the technology team and I was what we call a technology analyst, but really a network engineer.

And you started out really doing remote access, remote connectivity with technologies such as ISDN.

That was kind of the basis which we used back then. But I quickly learned the fact that the network is the business.

And if you think about an investment bank, and this still holds true today, it's something that's still core, I think, to when I tell the story and share the story maybe with clients and others, is that in a bank, if you don't have connectivity and the network goes down, there's nobody to trade with because you can't trade with yourself.

So you need, the network is the heart of everything that sits in the financial services industry, at least that's my opinion.

And really that ability to connect with other institutions, with stock exchanges or Borsa or whatever it might be, the ability to consume information in terms of market data, that network is fundamental in terms of the ability for the bank to operate.

And the reputational risk that that carries in terms of the network goes down and the institution is out of the market, that carries both financial penalties, compliance risk, but also a reputational risk that's hard to restore.

And as a young engineer, I look back in that and I think, wow, that was a lot of responsibility to carry.

And I think I'm very grateful for the managers and the team that I had around me.

They were prepared to put the confidence in such a young person to make the changes that we used to do.

We would be evolving the Internet and evolving the market data connectivity.

And to give you an idea, just a sense of where we were, I remember the Internet access for the entire organisation in London, which was still quite a big organisation at the time, was two megabits per second, replicating over two locations.

So our aggregate bandwidth to the Internet was a huge four megabits per second, if you can imagine.

And yeah, at that point, things just grew. I mean, we began to have to deal with things like the Y2K issues that came up.

So for those that were around in those days, that was an interesting times.

And then of course, we began to build further and further on that and began to learn about things like firewalls, proxies, all of the stuff that today makes up many of the services that we sell, we provide at Cloudflare as well.

And yeah, could you double click on that, especially from a financial services mindset, Y2K for anyone on the outside was simply like, okay, yep, the computers won't be able to understand time.

And you really didn't have any application to apply that to or apply that fear to.

But I assume, being in the heart of the financial technical industry, that bug actually had repercussions, like where you could see like, oh, if all of a sudden these files don't process this appropriately, we're in a heck of a mess.

Like, how did that look like on the front lines?

Yeah, I mean, firstly, I had very little to compare against, so it was kind of one of my first experience.

I remember arriving and speaking with my manager at the time, he said, would you mind coming in at New Year's Eve?

So the entire team celebrated the new millennium. And you can't leave until it's fixed, or the apocalypse happens, like one or the other?

Pretty much.

And you know what, it was the most underwhelming experience you can imagine, because we spent months basically running tests, tools, writing scripts, going through the engineering mindset of, let's model it in a lab, let's make sure that we accelerate, see what happens.

And I think for me, it drove up that real belief into an engineering mindset of, we test the hypothesis, we rely on data to make a decision.

And ultimately, we got to the day and the evening, we ran all the tests, the clock struck 12, we ran the test again, and we were done.

And everybody was looking at each other thinking, was that it?

And we kind of laughed and said, okay, well, let's make sure when the US wakes up.

And I remember walking out into the street as soon as it was around the dawn of the 1st of January 2000.

And obviously, all of the revelers had had their party, London, you know, people had been celebrating the fireworks, the place was a mess.

And we came out, we thought, well, that was the new millennium, and let's go.

And, you know, I look back at it now, and I think, you know, maybe it was a much ado about nothing.

But look, I mean, it taught us a couple of important lessons, which was to have that foresight about what comes next to try and anticipate problems that might come up in the future.

But as I said, that core engineering mindset of let's test, let's validate, let's get data, let's make sure that we take all of the steps necessary so that when we make a decision, and when we get into production, we do so in a safe and reliable way.

And using that as a segue, you know, thinking of the future, how did, how did that bring you to your career at Klaffler?

How did, you know, what drove or interested you about the work that Klaffler was doing?

Yeah. So I think over the years, one of the things that's always been coached to me is whatever you do, put the customer or put your key stakeholder at the center of the problem.

And anything you do after that will probably come out okay. So whether it be when I was at the bank, and maybe, you know, dealing with the mechanics of, you know, connectivity and dealing with the internal teams there, or whether it be, you know, part of my sales career as a systems engineer or solution engineer, put the customer or at the center of what you do.

And I think about today at Klaffler, if we think about the clients that we work with, and we think about the problems that they have to solve, then typically they have a couple of demographics to deal with.

They've got their employees, and trying to provide connectivity so that people can reach the applications they need to consume to do the work that they do.

But also they need to serve their clients. So you've got the general Internet population that might be consuming the services that we're providing as a digital channel or whatever it might be.

So if you think about it, the needs of the network and ability to provide a network or a platform to serve the needs of those two demographics, employees and customers.

Klaffler is uniquely positioned because we actually have a unified network that can solve the problems that exist around those populations, whether it be around remote connectivity to support hybrid work, whether it be around providing content delivery networking so that we can enable global coverage and access to new markets so that people can consume and use the services that, you know, our clients are trying to provide to their customers.

Klaffler is uniquely positioned because we have that platform, that global network in 300 locations, and we can basically solve the problems that exist around those domains, whether it be around dealing with the adverse conditions we see on the Internet with security controls, whether it be around providing, you know, the capacity and the bandwidth to meet peak demand and things like Black Friday or peak season trading, whether it be around providing new capabilities around emerging edge compute use cases.

So when we begin to think about modern AI use cases and begin to use corpus of data to train new models, and we think about having to serve those at the edge so that we get fast response times, we have the ability to operate at scale.

Again, you know, I believe Klaffler is uniquely positioned.

So the lessons I learned maybe from those early days about putting the customer first and putting the customer at the center of what we do, and again, thinking a little bit about the core networking proposition, the fact that the network really is the business, you know, Klaffler embodies that, and that's the reason why, you know, it attracted me as a company that I wanted to work for, and I'm very proud of the fact that I've got a team that kind of now focuses on those problems with our clients.

So ultimately you were just as attracted to the network as you were, you know, back in the day when you were learning about networking.

But you know what, that day that, you know, you first plugged in the token ring or the Internet connection, it stays with you, and today, you know, I'm still as passionate about networking, still interested about networking, but of course networking has become so much more, you know, I think it's easy to think about moving packets, but the reality is that networking has become the enabler for much of what we do.

If you think about the fact that that first computer I had, it did actually have a hard drive, it had a floppy disk as well, but it had a CPU, and if you think about it, those two things were tightly coupled.

Today, if you think about modern computing architectures and distributed systems, we took that CPU resource and we took the storage and we pulled it apart, but the thing is, is that the thing that sits in the middle is the network, that's the thing that enables these two parts to communicate.

I mean, you need the data, you need to store it somewhere, you need to process it somewhere, and the fact that you have this low latency, highly reliable, performant interconnect that sits in the middle, that is both secure and resilient, and provides the privacy and the confidentiality of that communication flow, that thing that sits in the middle, and it might be today that computer and that storage are, you know, many hundreds of miles apart, that thing still fascinates me, because ultimately it's the thing that makes all of this other stuff work, and used to be a bit of a running joke, but if ever there was a problem, oh, it must be a networking issue, start with the network and build up.

Well, you know, that's something that, first of all, I took offense at it, and I thought that's a bit painful, but now I realize actually it's a compliment, because it goes to demonstrate just the importance that the networking services and the security capabilities that it provides entails in terms of being able to support everything that comes about, and whether that be around data analytics, whether it be around, you know, artificial intelligence and the whole world we're moved to in terms of being able to do large language models, whether it be around systems of record, like SAP, all of that stuff is dependent on networking connectivity, and yeah, I'm very fortunate to have been part of that journey.

Awesome, and with our final minute, you know, what are you most excited about for the future of the Internet, or for some of the work that Cloudflare's doing?

Sure, I mean, it comes back down to, we have this platform, and for sure we can solve many problems with it today, but I think there's a whole class of problem, a whole set of domains that yet we haven't yet uncovered or touched, and I think there's some really smart and intelligent people who are thinking a little bit about how we solve those problems, whether it be around the issues around privacy, whether it be around the ability to operate at scale, whether it be around managing things like sovereignty and the challenges that go around the way that networks are built today and the geopolitical issues, whether it be around dealing with both the benefit and the promise of artificial intelligence, but also dealing with the complexities it brings as well.

All of those things, I think, are problem domains that I think we are well served, or well placed to serve, and yeah, I mean, it's interesting to see what comes next.

I mean, many years ago, I might have said, we're done with networking, it's just about connecting a couple of computers together and we're finished, but it has continued to be the enabler and the catalyst to many of the future developments I've saw in my career, and I'm sure beyond my career and into the long term, those will continue to happen.

So yeah, interested to see what comes next.

Awesome. Well, Andy, thank you for taking the time to chat with me. It was great going from early days and early text gaming to the future of the Internet with you.

Fantastic. And thanks for having us on. And yeah, pleasure to have the opportunity to speak today.

Awesome. And for everyone watching, either the live version or one of the reruns, thank you for joining us and until next time.