WEBVTT 00:00:00.000 --> 00:00:03.160 You know, if you've ever had that perfect experience 00:00:03.160 --> 00:00:05.519 online. Oh, yeah. I know exactly what you mean. 00:00:05.599 --> 00:00:07.339 Where you're watching a super high -res video 00:00:07.339 --> 00:00:11.119 and it just never buffers. Or you download this 00:00:11.119 --> 00:00:14.439 massive gigabyte -sized application update in 00:00:14.439 --> 00:00:17.379 what feels like seconds. Or, and this is the 00:00:17.379 --> 00:00:19.339 big one for me, you're trying to buy something 00:00:19.339 --> 00:00:22.120 during a huge online sale and the site actually 00:00:22.120 --> 00:00:25.510 works. It doesn't time out. Exactly. When that 00:00:25.510 --> 00:00:27.750 happens, you've just experienced an infrastructure 00:00:27.750 --> 00:00:30.350 that is working so well, it's basically invisible. 00:00:30.769 --> 00:00:32.289 And that's what we're talking about today, isn't 00:00:32.289 --> 00:00:34.829 it? The invisible backbone of the Internet. Right. 00:00:34.950 --> 00:00:37.490 Today, we are pulling back the curtain on the 00:00:37.490 --> 00:00:41.469 company that, well, quietly powers a huge amount 00:00:41.469 --> 00:00:44.409 of that modern web experience, Akamai Technologies. 00:00:44.549 --> 00:00:47.049 It really is the perfect case study. If you, 00:00:47.189 --> 00:00:49.130 the listener, are looking for that short pet 00:00:49.130 --> 00:00:51.590 to really understanding how this complex digital 00:00:51.590 --> 00:00:54.659 world works, Akamai. is where you start. And 00:00:54.659 --> 00:00:57.600 the name itself is a hint, right? It is. The 00:00:57.600 --> 00:01:00.899 name Akamai means clever or, you know, in a more 00:01:00.899 --> 00:01:04.140 modern sense, cool in Hawaiian. And that really 00:01:04.140 --> 00:01:06.719 points to the sheer mathematical sophistication 00:01:06.719 --> 00:01:08.599 that's under the hood of everything they do. 00:01:08.939 --> 00:01:11.000 Okay, so let's set our mission for this deep 00:01:11.000 --> 00:01:14.459 dive. We want to unpack Akamai's massive evolution. 00:01:14.920 --> 00:01:16.799 Yeah, because they didn't just stay in one lane. 00:01:16.939 --> 00:01:19.359 They started as a content delivery network, a 00:01:19.359 --> 00:01:22.859 CDN, just focused on making things faster. But 00:01:22.859 --> 00:01:25.260 they've made this incredibly calculated, you 00:01:25.260 --> 00:01:28.219 could even say aggressive strategic shift. Now 00:01:28.219 --> 00:01:30.900 they're a leader in cybersecurity and more recently, 00:01:31.019 --> 00:01:33.760 a major player in cloud services. So we're going 00:01:33.760 --> 00:01:36.079 to dissect their technical DNA. We'll track their 00:01:36.079 --> 00:01:38.040 growth through some massive acquisitions, which 00:01:38.040 --> 00:01:40.299 is a whole story in itself. And we'll look at 00:01:40.299 --> 00:01:42.840 the sometimes controversial place they hold in 00:01:42.840 --> 00:01:44.599 the history of the global network. Before we 00:01:44.599 --> 00:01:46.799 jump in, let's just ground everyone. What is 00:01:46.799 --> 00:01:50.840 Akamai today? OK, so at its core. Akamai is an 00:01:50.840 --> 00:01:52.939 American computer networking company. They're 00:01:52.939 --> 00:01:55.180 headquartered in Cambridge, Massachusetts. And 00:01:55.180 --> 00:01:56.879 their specialties now cover that whole stack. 00:01:57.560 --> 00:02:01.040 CDN, cybersecurity, DDoS mitigation, and a really 00:02:01.040 --> 00:02:03.319 fast -growing suite of cloud computing services. 00:02:03.680 --> 00:02:05.620 And just to understand the scale we're dealing 00:02:05.620 --> 00:02:08.460 with, they're a component of the S &P 500. Which 00:02:08.460 --> 00:02:10.120 tells you right away they're not just a tech 00:02:10.120 --> 00:02:12.699 company. They are considered systemically important 00:02:12.699 --> 00:02:14.939 to the U .S. economy. They're stable. They're 00:02:14.939 --> 00:02:18.060 foundational. And the numbers back that up completely. 00:02:18.280 --> 00:02:21.500 I mean, looking at their 2024 figures, this is 00:02:21.500 --> 00:02:24.300 a giant. We're talking about U .S. $3 .99 billion 00:02:24.300 --> 00:02:29.060 in revenue. With a net income of U .S. $505 million. 00:02:29.699 --> 00:02:32.900 And they do all this with over 10 ,700 employees 00:02:32.900 --> 00:02:35.490 around the world. So the takeaway is clear. This 00:02:35.490 --> 00:02:38.150 company has evolved way beyond its dot -com origins. 00:02:38.289 --> 00:02:40.669 It's now, you know, critical global infrastructure. 00:02:41.030 --> 00:02:42.750 So let's go all the way back to the beginning. 00:02:42.909 --> 00:02:45.830 Let's start at the genesis, 1998, because this 00:02:45.830 --> 00:02:47.650 company didn't come from a business school pitch 00:02:47.650 --> 00:02:50.439 deck. It started as a math problem. It did. It 00:02:50.439 --> 00:02:53.159 started as a very abstract, very difficult mathematical 00:02:53.159 --> 00:02:56.039 problem in MIT designed to fix something that 00:02:56.039 --> 00:02:58.080 was fundamentally broken about the early web. 00:02:58.240 --> 00:02:59.699 And that's really the key difference here. The 00:02:59.699 --> 00:03:02.699 whole intellectual origin story starts with the 00:03:02.699 --> 00:03:04.800 co -founders, Daniel Lewin and F. Thompson Leighton. 00:03:04.919 --> 00:03:06.939 Right. And their research was all about tackling 00:03:06.939 --> 00:03:09.620 this issue of Internet hotspots. Explain that. 00:03:09.680 --> 00:03:12.120 What was a hotspot back then? So imagine it's 00:03:12.120 --> 00:03:14.740 the late 90s. Something goes viral, a big news 00:03:14.740 --> 00:03:17.569 event, maybe one of the first viral videos. The 00:03:17.569 --> 00:03:19.569 sheer number of people trying to access that 00:03:19.569 --> 00:03:22.409 one piece of content from its single origin server 00:03:22.409 --> 00:03:25.210 would just overwhelm it. The server just falls 00:03:25.210 --> 00:03:27.750 over. It falls over. The internet was incredibly 00:03:27.750 --> 00:03:30.110 fragile. The system would crash. Performance 00:03:30.110 --> 00:03:32.090 would be terrible for everyone. It just couldn't 00:03:32.090 --> 00:03:36.930 scale. So how did they solve what seems like, 00:03:36.969 --> 00:03:39.289 I mean, an insurmountable technical problem? 00:03:39.490 --> 00:03:41.250 It was all based on a concept they developed 00:03:41.250 --> 00:03:43.840 called consistent hashing. Okay. You're going 00:03:43.840 --> 00:03:45.419 to have to break that down for us. Right. So 00:03:45.419 --> 00:03:47.360 think of the Internet as this, I don't know, 00:03:47.419 --> 00:03:50.240 a giant library. And when a book becomes super 00:03:50.240 --> 00:03:53.080 popular, everyone rushes to that one single shelf 00:03:53.080 --> 00:03:55.740 to get it. And the shelf breaks. The shelf breaks. 00:03:55.780 --> 00:03:58.719 The whole library grinds to a halt. What consistent 00:03:58.719 --> 00:04:03.819 hashing does is it provides this really elegant 00:04:03.819 --> 00:04:06.860 mathematical way to put copies of that book. 00:04:07.289 --> 00:04:09.550 On thousands of different shelves all over the 00:04:09.550 --> 00:04:12.310 library in a distributed network. So you're not 00:04:12.310 --> 00:04:15.389 all crowding the same spot. You're not. But here's 00:04:15.389 --> 00:04:17.790 the really clever part, the consistent part of 00:04:17.790 --> 00:04:20.829 the name. If you need to add a new shelf or take 00:04:20.829 --> 00:04:23.189 one away for maintenance, the math behind it 00:04:23.189 --> 00:04:26.230 ensures you only have to move a tiny, minimal 00:04:26.230 --> 00:04:29.110 number of books. Ah, so it's efficient. You don't 00:04:29.110 --> 00:04:31.329 have to reorganize the entire library. Exactly. 00:04:31.810 --> 00:04:34.290 Dramatically improved efficiency, scalability, 00:04:34.569 --> 00:04:38.410 and crucially, stability. It's really the concept 00:04:38.410 --> 00:04:40.750 that allowed the modern high -traffic Internet 00:04:40.750 --> 00:04:43.689 to even function. And that conceptual breakthrough, 00:04:44.009 --> 00:04:46.290 it wasn't just theory. It gave them a viable 00:04:46.290 --> 00:04:48.410 product almost right away. It was immediate. 00:04:48.750 --> 00:04:51.470 Their business proposition was so powerful that 00:04:51.470 --> 00:04:55.110 it was chosen as a finalist in the 1998 MIT $50K 00:04:55.110 --> 00:04:58.350 competition, which is a huge deal. That validation 00:04:58.350 --> 00:05:00.329 was the springboard. So they incorporated pretty 00:05:00.329 --> 00:05:02.410 quickly after that. Incredibly quickly. They 00:05:02.410 --> 00:05:04.430 brought in Jonathan Selig and Randall Kaplan 00:05:04.430 --> 00:05:06.490 to help with the business side, and they officially 00:05:06.490 --> 00:05:10.410 incorporated on August 20, 1998. That kind of 00:05:10.410 --> 00:05:13.050 rapid success, especially then, it must have 00:05:13.050 --> 00:05:15.250 meant they needed to bring in some serious business 00:05:15.250 --> 00:05:17.829 veterans to manage the growth. Oh, for sure. 00:05:17.910 --> 00:05:20.490 You can't scale from an academic project to a 00:05:20.490 --> 00:05:23.290 global company without that. They recruited two 00:05:23.290 --> 00:05:26.540 key people, George Conrads, And Paul Sagan. Icon 00:05:26.540 --> 00:05:29.000 Res was a big name. A huge name. He'd been the 00:05:29.000 --> 00:05:32.180 chairman of BBN Corp and a senior VP of U .S. 00:05:32.199 --> 00:05:34.879 operations for IBM. He brought that, you know, 00:05:34.899 --> 00:05:37.720 that IBM level corporate seriousness. And Sagan, 00:05:37.860 --> 00:05:40.319 he was a former president of New Media for Time, 00:05:40.360 --> 00:05:43.129 Inc. So he understood the content side. That's 00:05:43.129 --> 00:05:46.089 a powerful combination, IBM Gravitas and new 00:05:46.089 --> 00:05:49.009 media savvy. It was essential. And their timeline 00:05:49.009 --> 00:05:52.889 from zero to the NASDAQ is just a perfect picture 00:05:52.889 --> 00:05:55.470 of that late 90s tech boom. It's wild. It's an 00:05:55.470 --> 00:05:57.990 absolutely explosive trajectory. So they bring 00:05:57.990 --> 00:06:01.589 Conrad's on as CEO in April 1999. They launched 00:06:01.589 --> 00:06:03.389 their commercial service that very same month. 00:06:03.490 --> 00:06:05.829 Okay. And then just six months later, on October 00:06:05.829 --> 00:06:09.350 29, 1999, they were listed on the NASDAQ. Wow. 00:06:09.509 --> 00:06:12.449 From an MIT research paper to a publicly traded 00:06:12.449 --> 00:06:14.769 company in just over a year. You just don't see 00:06:14.769 --> 00:06:16.790 speed like that very often. It's remarkable. 00:06:17.069 --> 00:06:19.850 And it's here in the middle of this incredible 00:06:19.850 --> 00:06:23.550 success story that the narrative takes just a 00:06:23.550 --> 00:06:26.910 profoundly tragic turn. We really have to address 00:06:26.910 --> 00:06:30.430 the event that changed the company and the world 00:06:30.430 --> 00:06:34.870 forever. Yes. The story is, it's just shadowed 00:06:34.870 --> 00:06:37.709 by the tragedy of September 11, 2001. The co 00:06:37.709 --> 00:06:40.689 -founder, Daniel Lewin. Daniel M. Lewin. He was 00:06:40.689 --> 00:06:43.810 killed aboard American Airlines Flight 11. That 00:06:43.810 --> 00:06:45.810 was the first plane to hit the World Trade Center. 00:06:46.310 --> 00:06:49.149 Lewin was a genius, not just in computer science, 00:06:49.230 --> 00:06:51.250 but he was also a former officer in the Israeli 00:06:51.250 --> 00:06:53.930 military. And he was seated close to the hijackers? 00:06:54.129 --> 00:06:56.990 He was. And based on reports and analysis, it's 00:06:56.990 --> 00:06:59.060 believed he may have tried to stop them. He was 00:06:59.060 --> 00:07:01.220 likely the first victim of the attacks, stabbed 00:07:01.220 --> 00:07:03.240 by one of the terrorists. That's just an immense, 00:07:03.420 --> 00:07:05.620 an unbelievable loss, not just for the company, 00:07:05.639 --> 00:07:07.579 but for the entire field of Internet research. 00:07:07.759 --> 00:07:10.680 A visionary founder, gone. And for the company, 00:07:10.920 --> 00:07:13.660 still in its early days, that's a blow that could 00:07:13.660 --> 00:07:15.439 have been fatal. How did they manage to move 00:07:15.439 --> 00:07:18.040 forward? It was a defining moment for their culture 00:07:18.040 --> 00:07:20.699 and their leadership. George Conrads really had 00:07:20.699 --> 00:07:22.899 to steer the company through that immediate aftermath 00:07:22.899 --> 00:07:25.660 and then right into the dot -com bust that followed. 00:07:25.959 --> 00:07:28.920 He stayed on until 2005. And that's when Paul 00:07:28.920 --> 00:07:31.819 Sagan, the media expert, took over. That's right. 00:07:31.879 --> 00:07:34.899 Sagan took the reins as CEO. And his tenure seems 00:07:34.899 --> 00:07:37.240 to be about proving Akamai could be more than 00:07:37.240 --> 00:07:39.500 just a one -trick pony, that it was a sustainable 00:07:39.500 --> 00:07:41.980 business. That's exactly it. Sagan recognized 00:07:41.980 --> 00:07:46.079 that being only a CDN made them vulnerable. Competitors 00:07:46.079 --> 00:07:49.040 could pop up. Prices could drop. So his strategy 00:07:49.040 --> 00:07:52.060 was all about aggressive diversification. Expanding 00:07:52.060 --> 00:07:55.920 the services beyond just caching content. Precisely. 00:07:56.000 --> 00:07:58.560 And it worked. Under his leadership, the company 00:07:58.560 --> 00:08:01.120 grew enormously. By the time he left the CEO 00:08:01.120 --> 00:08:04.279 role, they were up to $1 .37 billion in revenue. 00:08:04.540 --> 00:08:06.800 He really built the foundation for the integrated 00:08:06.800 --> 00:08:09.040 platform they are today. And then the next big 00:08:09.040 --> 00:08:12.459 leadership shift. In 2013, it was a return to 00:08:12.459 --> 00:08:15.259 the company's deepest roots. It was. In 2013, 00:08:15.620 --> 00:08:17.920 co -founder Tom Layton, the mathematician, the 00:08:17.920 --> 00:08:19.939 guy behind the original consistent hashing research, 00:08:20.199 --> 00:08:23.120 he was elected CEO, replacing Sagan. What did 00:08:23.120 --> 00:08:25.879 that signal to the market? It signaled a strategic 00:08:25.879 --> 00:08:29.620 doubling down on technology and innovation. It 00:08:29.620 --> 00:08:31.579 was a statement that in a world that was getting 00:08:31.579 --> 00:08:35.240 more complex, Akamai's edge would come from deep 00:08:35.240 --> 00:08:37.639 engineering, not just from business strategy. 00:08:37.940 --> 00:08:40.659 And throughout all this, the market really recognized 00:08:40.659 --> 00:08:42.769 their importance. Absolutely. They were getting 00:08:42.769 --> 00:08:45.269 constant validation. They were added to the Russell 00:08:45.269 --> 00:08:48.429 Indexes in 2001. And then the really big one, 00:08:48.529 --> 00:08:52.009 they joined the S &P 500 Index in July of 2007. 00:08:52.389 --> 00:08:54.429 Which means you're no longer just a successful 00:08:54.429 --> 00:08:56.450 tech company. You're a pillar. You are a pillar 00:08:56.450 --> 00:08:58.830 of the American industrial landscape. And their 00:08:58.830 --> 00:09:00.830 physical presence, their headquarters kind of 00:09:00.830 --> 00:09:03.690 grew along with their stature. It did. It mirrored 00:09:03.690 --> 00:09:05.779 their journey. They started in Technology Square, 00:09:05.960 --> 00:09:07.879 then they were spread out across multiple buildings 00:09:07.879 --> 00:09:10.799 in Cambridge Center. And this all culminated 00:09:10.799 --> 00:09:13.980 in December 2019. With the new building. A massive 00:09:13.980 --> 00:09:16.559 purpose -built headquarters at 145 Broadway, 00:09:16.940 --> 00:09:20.259 right in the heart of Kendall Square. And that's 00:09:20.259 --> 00:09:22.539 not just a real estate move. It's symbolic. It 00:09:22.539 --> 00:09:25.740 plants Akamai's flag right next to MIT in the 00:09:25.740 --> 00:09:27.539 center of one of the world's biggest innovation 00:09:27.539 --> 00:09:30.879 hubs. Okay, let's shift gears now and talk about 00:09:30.879 --> 00:09:33.580 the actual magic behind the curtain. The core 00:09:33.580 --> 00:09:35.960 of their entire operation is this thing they 00:09:35.960 --> 00:09:39.080 call the Akamai Intelligent Edge Platform. Right. 00:09:39.159 --> 00:09:41.759 We hear words like cloud and edge thrown around 00:09:41.759 --> 00:09:44.840 constantly, but what does Intelligent Edge really 00:09:44.840 --> 00:09:47.580 mean conceptually for someone trying to get their 00:09:47.580 --> 00:09:49.899 head around this stuff? Okay, so the Intelligent 00:09:49.899 --> 00:09:52.840 Edge Platform is first and foremost a massively 00:09:52.840 --> 00:09:55.379 distributed cloud computing platform. That's 00:09:55.379 --> 00:09:57.600 the key. Massively distributed. So what's the 00:09:57.600 --> 00:09:59.679 difference between this and, say, a traditional 00:09:59.679 --> 00:10:02.519 cloud provider like AWS? The difference is its 00:10:02.519 --> 00:10:07.299 scale and its topology, its shape. When we say 00:10:07.299 --> 00:10:09.840 edge, we mean they have physically placed their 00:10:09.840 --> 00:10:12.460 servers as close as humanly possible to you, 00:10:12.500 --> 00:10:14.840 the user. It's at the final mile of the Internet. 00:10:14.940 --> 00:10:17.039 That's totally different from AWS or Azure, who 00:10:17.039 --> 00:10:19.759 rely on these huge centralized mega regions. 00:10:19.980 --> 00:10:22.779 And this scale, this distribution. That's really 00:10:22.779 --> 00:10:24.539 their competitive advantage, isn't it? Let's 00:10:24.539 --> 00:10:26.799 try to quantify this edge. Just how big are we 00:10:26.799 --> 00:10:29.360 talking? The scale is, frankly, astronomical. 00:10:29.659 --> 00:10:31.779 It's hard to even comprehend. The Akamai network 00:10:31.779 --> 00:10:35.539 today consists of approximately 365 ,000 servers. 00:10:35.799 --> 00:10:39.139 Wow. 365 ,000 individual points of presence. 00:10:39.320 --> 00:10:41.600 And they are meticulously spread across more 00:10:41.600 --> 00:10:43.919 than 135 countries and territories. So they're 00:10:43.919 --> 00:10:46.159 everywhere. They're everywhere. And they sit 00:10:46.159 --> 00:10:49.279 on roughly 1 ,350 of the world's different networks. 00:10:49.870 --> 00:10:52.269 So no matter where you're connecting from, the 00:10:52.269 --> 00:10:55.450 content, the data, it's never far away. That's 00:10:55.450 --> 00:10:57.230 incredible. But just having a lot of servers 00:10:57.230 --> 00:11:00.169 doesn't make it intelligent. What's the smart 00:11:00.169 --> 00:11:03.889 part? The intelligence is the software. It's 00:11:03.889 --> 00:11:06.769 the real -time data collection and the complex 00:11:06.769 --> 00:11:09.169 algorithmic brain that runs on top of all those 00:11:09.169 --> 00:11:11.490 servers. The network is constantly monitoring 00:11:11.490 --> 00:11:13.610 itself and the wider Internet. What's it looking 00:11:13.610 --> 00:11:15.789 for? It's gathering information on global traffic. 00:11:15.870 --> 00:11:17.889 It's identifying congestion points. It's tracking 00:11:17.889 --> 00:11:20.350 trouble spots on the network. And it's calculating 00:11:20.350 --> 00:11:23.230 the absolute best routes for data all in real 00:11:23.230 --> 00:11:26.230 time. This proprietary software, which all goes 00:11:26.230 --> 00:11:28.269 back to that original consistent hashing idea, 00:11:28.490 --> 00:11:30.990 it ensures the system doesn't just store content. 00:11:31.549 --> 00:11:34.470 It actively and intelligently routes you to the 00:11:34.470 --> 00:11:37.809 fastest, healthiest, least congested server available 00:11:37.809 --> 00:11:41.289 at that exact nanosecond. So it's not a static 00:11:41.289 --> 00:11:44.990 map. It's a living, breathing, globally distributed 00:11:44.990 --> 00:11:47.250 traffic controller. That's the perfect way to 00:11:47.250 --> 00:11:50.070 put it. It's a gigantic self -optimizing system. 00:11:50.610 --> 00:11:53.330 Think of it like this. If you're going on a road 00:11:53.330 --> 00:11:55.909 trip, a normal DNS server just tells you the 00:11:55.909 --> 00:11:58.070 address of your final destination. Right. Here's 00:11:58.070 --> 00:12:00.990 where you're going. Akamai's system is like Waze 00:12:00.990 --> 00:12:03.830 or Google Maps for the Internet. It's constantly 00:12:03.830 --> 00:12:06.889 updating you on accidents, on traffic jams, and 00:12:06.889 --> 00:12:09.750 it's diverting you in real time to the fastest 00:12:09.750 --> 00:12:12.230 alternate route so you get to the nearest best 00:12:12.230 --> 00:12:14.750 exit the edge server with the absolute minimum 00:12:14.750 --> 00:12:16.690 delay. Okay, so let's walk through an actual 00:12:16.690 --> 00:12:18.669 example. Say I click on a link to download a 00:12:18.669 --> 00:12:21.470 huge software update from a big company. How 00:12:21.470 --> 00:12:24.129 does Akamai step in and do its thing? How does 00:12:24.129 --> 00:12:26.610 it avoid the middle mile? Okay, so it's basically 00:12:26.610 --> 00:12:28.710 a four -step process that hijacks the standard 00:12:28.710 --> 00:12:31.690 web request to make it better. Step one, you 00:12:31.690 --> 00:12:33.899 click the link. you initiate the URL request. 00:12:34.139 --> 00:12:36.980 Simple enough. Step two, that triggers a DNS 00:12:36.980 --> 00:12:40.940 request. But, and this is the crucial part, instead 00:12:40.940 --> 00:12:43.240 of going to the software company's big, faraway 00:12:43.240 --> 00:12:46.559 data center, the request is intercepted and redirected 00:12:46.559 --> 00:12:49.700 to Akamai's own authoritative DNS servers. This 00:12:49.700 --> 00:12:52.600 is the moment Akamai takes control. So my request 00:12:52.600 --> 00:12:54.820 for directions is handed off to Akamai's Waze. 00:12:54.840 --> 00:12:57.980 What's step three? In step three, Akamai's mapping 00:12:57.980 --> 00:13:00.639 system that brain running all the algorithms, 00:13:00.799 --> 00:13:03.600 retrieves an IP address. But that IP address 00:13:03.600 --> 00:13:05.480 doesn't point to the origin server. It points 00:13:05.480 --> 00:13:08.240 to the single optimal Akamai Edge server that 00:13:08.240 --> 00:13:10.080 is closest to you and has the best connection 00:13:10.080 --> 00:13:13.139 at that moment. And step four. Step four. Your 00:13:13.139 --> 00:13:15.159 browser or device connects directly to that chosen 00:13:15.159 --> 00:13:17.639 Akamai Edge server, and the download starts immediately. 00:13:17.960 --> 00:13:20.100 So I'm not going across the country or across 00:13:20.100 --> 00:13:22.679 the ocean to get the file. I'm talking to a server 00:13:22.679 --> 00:13:25.279 that might be in my city. Maybe even my neighborhood. 00:13:25.539 --> 00:13:28.360 Exactly. But how does that server have the file 00:13:28.360 --> 00:13:30.500 already? That's the question. That's the concept 00:13:30.500 --> 00:13:34.059 of mirroring and caching. Akamai transparently 00:13:34.059 --> 00:13:36.720 mirrors all the key web content from its customers 00:13:36.720 --> 00:13:40.419 to HTML, JavaScript, but also massive software 00:13:40.419 --> 00:13:42.840 downloads, video streams, everything. The server 00:13:42.840 --> 00:13:45.059 selection process knows what you want and where 00:13:45.059 --> 00:13:47.279 you are, and it has it ready for you. And this 00:13:47.279 --> 00:13:49.419 all serves to bypass what you call the middle 00:13:49.419 --> 00:13:51.659 mile bottleneck. What's so bad about the middle 00:13:51.659 --> 00:13:53.879 mile? The middle mile is the long, unpredictable 00:13:53.879 --> 00:13:57.110 haul. It's the transoceanic fiber cables, the 00:13:57.110 --> 00:13:59.590 international pathways. It's where data gets 00:13:59.590 --> 00:14:02.090 stuck in traffic jams, where different networks 00:14:02.090 --> 00:14:04.610 don't talk to each other efficiently. It's slow 00:14:04.610 --> 00:14:07.429 and unreliable. So by delivering from the last 00:14:07.429 --> 00:14:10.460 mile. From that server near me. Akamai gives 00:14:10.460 --> 00:14:13.779 you faster downloads, way more reliability, and 00:14:13.779 --> 00:14:16.360 it allows for massive scalability during huge 00:14:16.360 --> 00:14:18.580 events. They have a whole product line, download 00:14:18.580 --> 00:14:20.860 delivery built just on this with detailed analytics 00:14:20.860 --> 00:14:23.399 so customers can see exactly how well it's working. 00:14:23.559 --> 00:14:25.559 The source material also mentioned they use other 00:14:25.559 --> 00:14:27.399 delivery methods sometimes beyond just their 00:14:27.399 --> 00:14:30.460 own servers. They do. They also use peer -to 00:14:30.460 --> 00:14:32.340 -peer networking for certain types of content. 00:14:32.620 --> 00:14:35.620 This means in some cases they can leverage the 00:14:35.620 --> 00:14:37.740 spare bandwidth of other end users' computers 00:14:37.740 --> 00:14:40.519 to help deliver content. It's just another layer 00:14:40.519 --> 00:14:42.960 of optimization. Okay, one last thing on the 00:14:42.960 --> 00:14:45.519 tech. In 2013, they launched something called 00:14:45.519 --> 00:14:48.940 the Open Initiative. Why did they need to open 00:14:48.940 --> 00:14:52.120 up what sounds like a very closed, very proprietary 00:14:52.120 --> 00:14:55.019 system? That was a really important strategic 00:14:55.019 --> 00:14:58.379 move towards becoming a platform. The web was 00:14:58.379 --> 00:15:00.799 getting more complex and big customers didn't 00:15:00.799 --> 00:15:03.759 just want a service, they wanted control. The 00:15:03.759 --> 00:15:06.539 open initiative gave customers and partners direct 00:15:06.539 --> 00:15:09.120 API access to the intelligent platform. So they 00:15:09.120 --> 00:15:10.879 could customize it. They could develop their 00:15:10.879 --> 00:15:12.720 own tools. They could integrate it deeply with 00:15:12.720 --> 00:15:14.580 their own DevOps pipelines. They could get real 00:15:14.580 --> 00:15:17.419 -time big data out of the system all through 00:15:17.419 --> 00:15:20.200 a single interface. It changed Akamai from being 00:15:20.200 --> 00:15:22.399 just a service provider to being programmable 00:15:22.399 --> 00:15:24.580 infrastructure. So Akamai's original mission 00:15:24.580 --> 00:15:27.080 was all about speed. But if you look at the last 00:15:27.080 --> 00:15:29.440 decade, the story is completely different. It's 00:15:29.440 --> 00:15:32.000 about fortification. It's a story driven almost 00:15:32.000 --> 00:15:35.899 entirely by acquisitions. A huge, highly deliberate, 00:15:36.019 --> 00:15:40.080 strategic pivot. They clearly recognized that 00:15:40.080 --> 00:15:42.519 just being fast isn't enough in the modern world. 00:15:42.659 --> 00:15:44.899 You have to be secure and you have to be able 00:15:44.899 --> 00:15:47.240 to run applications. And they made this official, 00:15:47.379 --> 00:15:49.320 didn't they? They reorganized the company. They 00:15:49.320 --> 00:15:52.809 did. On February 9, 2021, they announced this 00:15:52.809 --> 00:15:55.610 big reorganization into two distinct internal 00:15:55.610 --> 00:15:58.690 groups, security technology and edge technology. 00:15:59.049 --> 00:16:01.929 And they reestablished the CTO role, the chief 00:16:01.929 --> 00:16:05.049 technology officer role, naming Robert Blumhoff. 00:16:05.269 --> 00:16:08.190 That move alone tells you everything. It says 00:16:08.190 --> 00:16:10.250 that the next phase of growth is going to be 00:16:10.250 --> 00:16:13.090 driven by deep technology, not just by sales. 00:16:13.309 --> 00:16:15.269 So let's look at how they built that first pillar, 00:16:15.409 --> 00:16:17.450 the security technology group. It makes sense, 00:16:17.490 --> 00:16:19.220 right? If you're sitting at the edge of the Internet, 00:16:19.379 --> 00:16:21.299 you're the first line of defense. You have to 00:16:21.299 --> 00:16:23.440 be a shield. And the first thing they did was 00:16:23.440 --> 00:16:25.299 buy the muscle they needed to handle massive 00:16:25.299 --> 00:16:27.679 attacks. That was Prolexic. Prolexic Technologies. 00:16:28.840 --> 00:16:32.320 Acquired in February 2014 for about $390 million. 00:16:33.539 --> 00:16:36.159 Prolexic was a pure play cybersecurity provider 00:16:36.159 --> 00:16:40.559 that specialized in DDoS prevention. This wasn't 00:16:40.559 --> 00:16:42.539 just adding a feature. This immediately gave 00:16:42.539 --> 00:16:44.360 Akamai a world -class defensive architecture 00:16:44.360 --> 00:16:47.000 that could absorb and scrub some of the biggest 00:16:47.000 --> 00:16:49.440 attacks on the internet. So overnight, they became 00:16:49.440 --> 00:16:52.320 a frontline defender. Absolutely. But then the 00:16:52.320 --> 00:16:54.120 threat landscape changed. It wasn't just about 00:16:54.120 --> 00:16:56.039 big attacks from the outside anymore. It was 00:16:56.039 --> 00:16:58.139 about what happens when an attacker gets inside. 00:16:58.399 --> 00:17:01.580 Exactly. The idea of zero trust. And to get into 00:17:01.580 --> 00:17:04.259 that game, they made a massive acquisition. Guardicore. 00:17:04.380 --> 00:17:08.259 This one was huge. September 2021 for $600 million. 00:17:08.640 --> 00:17:11.579 A massive and deeply strategic purchase. This 00:17:11.579 --> 00:17:13.559 became the foundation for their whole Zero Trust 00:17:13.559 --> 00:17:15.980 security platform. OK, let's stop and define 00:17:15.980 --> 00:17:18.160 that for the listener. What is Zero Trust and 00:17:18.160 --> 00:17:20.720 what is segmentation? So traditional security 00:17:20.720 --> 00:17:23.240 is like a castle. It has a hard shell, a strong 00:17:23.240 --> 00:17:25.680 perimeter. But once you get inside, the interior 00:17:25.680 --> 00:17:28.480 is soft. A hacker can move around freely. They 00:17:28.480 --> 00:17:31.059 can move laterally. Right. Zero Trust says we 00:17:31.059 --> 00:17:33.839 trust nothing, even inside the network. And the 00:17:33.839 --> 00:17:36.859 way you enforce that is with segmentation. You 00:17:36.859 --> 00:17:39.359 divide the internal network into tiny isolated 00:17:39.359 --> 00:17:43.079 zones. So if one part gets compromised, the attacker 00:17:43.079 --> 00:17:45.579 is trapped. They're stuck in that one tiny segment 00:17:45.579 --> 00:17:48.859 and can't move laterally to steal critical data 00:17:48.859 --> 00:17:52.460 or infect other systems. GuardiCore gave Akamai 00:17:52.460 --> 00:17:54.839 the software to do that at a massive scale. So 00:17:54.839 --> 00:17:57.000 they've secured the perimeter with Prolexic. 00:17:57.000 --> 00:17:59.420 They've secured the inside with GuardiCore. What 00:17:59.420 --> 00:18:01.680 was next? The next major attack surface. Yeah. 00:18:01.759 --> 00:18:03.759 How do modern applications talk to each other? 00:18:03.940 --> 00:18:07.059 Through APIs, application programming interfaces. 00:18:07.500 --> 00:18:10.299 Right, which are everywhere now. And often very 00:18:10.299 --> 00:18:13.680 insecure. So in June of 2024, they bought No 00:18:13.680 --> 00:18:17.000 Name Security, an API security vendor, for $450 00:18:17.000 --> 00:18:19.660 million. This shows they're committed to covering 00:18:19.660 --> 00:18:21.799 every new vulnerability that emerges. And then 00:18:21.799 --> 00:18:23.480 they filled in the gaps with a bunch of other 00:18:23.480 --> 00:18:25.500 specialized firms. It was like a mosaic. They 00:18:25.500 --> 00:18:27.019 just systematically bought the pieces they needed. 00:18:27.180 --> 00:18:30.319 In December 2016, they bought CyberFend for bot 00:18:30.319 --> 00:18:32.500 detection, which is crucial for stopping credential 00:18:32.500 --> 00:18:35.460 stuffing and protecting e -commerce sites. They 00:18:35.460 --> 00:18:38.599 bought Soha Systems in October 2016 for secure 00:18:38.599 --> 00:18:41.779 enterprise access. They got secure web gateway 00:18:41.779 --> 00:18:45.160 tech by buying blocks in November 2015. They 00:18:45.160 --> 00:18:47.059 even bought a company called Inverse Inc. in 00:18:47.059 --> 00:18:49.640 February 2021 for its open source network access 00:18:49.640 --> 00:18:52.759 controller, PacketFence. They were building a 00:18:52.759 --> 00:18:55.059 complete end -to -end security stack. OK, so 00:18:55.059 --> 00:18:57.319 that's the security pillar. Let's pivot to the 00:18:57.319 --> 00:19:00.279 second massive expansion, the push into cloud 00:19:00.279 --> 00:19:02.619 computing. This feels like their biggest bet 00:19:02.619 --> 00:19:05.160 yet. It is arguably their most high stakes move. 00:19:05.420 --> 00:19:07.359 They didn't just want to deliver content anymore. 00:19:07.440 --> 00:19:09.769 They wanted to. host and run the applications 00:19:09.769 --> 00:19:12.009 themselves. They wanted to compete directly with 00:19:12.009 --> 00:19:14.029 the big hyperscalers. And the foundational move 00:19:14.029 --> 00:19:16.569 here was buying Linode. Linode. Yeah. February 00:19:16.569 --> 00:19:20.910 2022 for a whopping $900 million. Linode was 00:19:20.910 --> 00:19:23.509 the crucial missing piece. It gave them a strong 00:19:23.509 --> 00:19:26.009 existing cloud infrastructure service compute 00:19:26.009 --> 00:19:29.609 storage with a loyal developer -friendly community 00:19:29.609 --> 00:19:32.130 already built in. Why was buying Linode so essential? 00:19:32.230 --> 00:19:33.349 Couldn't they have just built that themselves? 00:19:33.829 --> 00:19:36.250 They could have, but it would have taken years 00:19:36.250 --> 00:19:38.940 and been incredibly expensive. Building a developer 00:19:38.940 --> 00:19:42.160 -friendly cloud platform from scratch is a monumental 00:19:42.160 --> 00:19:45.039 task. Linode gave them instant infrastructure 00:19:45.039 --> 00:19:47.759 and, just as importantly, instant credibility 00:19:47.759 --> 00:19:50.680 with developers. And the strategy was to merge 00:19:50.680 --> 00:19:53.519 Linode's compute with Akamai's Edge network. 00:19:53.880 --> 00:19:56.940 That's the secret sauce. By integrating Linode's 00:19:56.940 --> 00:20:00.119 compute layer with their existing 365 ,000 servers, 00:20:00.559 --> 00:20:03.640 Akamai instantly created a decentralized cloud 00:20:03.640 --> 00:20:06.240 offering. This is a strategic counter -argument 00:20:06.240 --> 00:20:08.940 to AWS or Google Cloud, where you have to pick 00:20:08.940 --> 00:20:12.400 a big centralized region to run your stuff. Akamai 00:20:12.400 --> 00:20:14.240 says, why not run it everywhere? And they got 00:20:14.240 --> 00:20:16.539 a huge piece of market validation for this strategy 00:20:16.539 --> 00:20:19.099 pretty quickly. A massive proof point. In February 00:20:19.099 --> 00:20:22.720 of 2025, Akamai announced this major win. They 00:20:22.720 --> 00:20:25.119 were chosen as a strategic cloud provider by 00:20:25.119 --> 00:20:26.940 one of the world's largest technology companies. 00:20:27.079 --> 00:20:29.720 And this wasn't a small deal. Not at all. It 00:20:29.720 --> 00:20:31.680 was a multi -year commitment to spend over $100 00:20:31.680 --> 00:20:34.259 million on their cloud services, the services 00:20:34.259 --> 00:20:37.150 built on that Linode foundation. When a massive 00:20:37.150 --> 00:20:39.309 tech peer gives you a nine -figure vote of confidence, 00:20:39.549 --> 00:20:41.950 it confirms your strategy is viable at the enterprise 00:20:41.950 --> 00:20:44.210 level. So before we move on, let's just quickly 00:20:44.210 --> 00:20:46.650 touch on the acquisitions that built the original 00:20:46.650 --> 00:20:48.730 CDN backbone we talked about earlier. Right. 00:20:48.769 --> 00:20:50.529 These were the deals that cemented their dominance 00:20:50.529 --> 00:20:53.529 back in the 2000s. They bought their rival. Speedera 00:20:53.529 --> 00:20:56.710 Networks in June 2005. They picked up another 00:20:56.710 --> 00:21:01.069 one, Netly Inc., in March 2007 for about $154 00:21:01.069 --> 00:21:04.069 million. And the Katendo acquisition was a big 00:21:04.069 --> 00:21:06.289 one, right? That was seen as a major move. A 00:21:06.289 --> 00:21:08.950 huge strategic move. They bought Katendo in March 00:21:08.950 --> 00:21:13.609 2012 for nearly $279 million. Katendo was known 00:21:13.609 --> 00:21:16.089 for really innovative mobile and dynamic content 00:21:16.089 --> 00:21:19.269 acceleration. Buying them helped Akamai stay 00:21:19.269 --> 00:21:21.920 ahead of the technology curve. And they also 00:21:21.920 --> 00:21:24.299 snapped up smaller specialized firms like Blaze 00:21:24.299 --> 00:21:26.740 Software and Fasasoft just to fine tune every 00:21:26.740 --> 00:21:28.920 last bit of performance out of the network. OK, 00:21:29.000 --> 00:21:31.619 so when a company sits at the very center of 00:21:31.619 --> 00:21:35.160 global data flow, when it's the pipe for securing, 00:21:35.559 --> 00:21:39.039 speeding up and now hosting web traffic, it's 00:21:39.039 --> 00:21:41.299 inevitable that it gets caught up in global politics 00:21:41.299 --> 00:21:44.259 and scrutiny. Absolutely. When you own that much 00:21:44.259 --> 00:21:46.680 of the global highway, you become a player in 00:21:46.680 --> 00:21:48.619 international affairs, whether you want to or 00:21:48.619 --> 00:21:51.549 not. So we have to look at some of the key historical 00:21:51.549 --> 00:21:54.269 incidents that put Akamai in the public spotlight. 00:21:54.809 --> 00:21:57.029 And we want to be really careful here to present 00:21:57.029 --> 00:22:00.009 the source material's claims impartially. A good 00:22:00.009 --> 00:22:02.369 place to start is the Al Jazeera incident back 00:22:02.369 --> 00:22:05.990 in 2003. Akamai was providing its content delivery 00:22:05.990 --> 00:22:08.690 services for the Arabic News Network during the 00:22:08.690 --> 00:22:10.910 first few weeks of the Iraq war. This was from 00:22:10.910 --> 00:22:14.670 March 28th to April 2nd, 2003. And then suddenly 00:22:14.670 --> 00:22:17.109 the service was cut off. It was terminated abruptly. 00:22:17.640 --> 00:22:20.000 And the claim at the time from the English language 00:22:20.000 --> 00:22:23.059 managing editor of Al Jazeera was that Akamai 00:22:23.059 --> 00:22:25.839 terminated the service due to political pressure. 00:22:25.980 --> 00:22:29.119 And the sources report that claim. They do. And 00:22:29.119 --> 00:22:31.779 it really highlights the incredibly difficult 00:22:31.779 --> 00:22:35.500 position Akamai is in. As a U .S.-based company, 00:22:35.660 --> 00:22:38.299 they face intense pressure over who they serve, 00:22:38.420 --> 00:22:40.779 especially when that client's content is seen 00:22:40.779 --> 00:22:42.839 as controversial by certain governments or groups. 00:22:43.309 --> 00:22:46.049 Moving on from geopolitics to, well, internal 00:22:46.049 --> 00:22:48.390 issues, Akamai also got caught up in some of 00:22:48.390 --> 00:22:51.630 the big ethical scandals of the financial world. 00:22:51.789 --> 00:22:53.410 That's the insider trading charge from 2013. 00:22:53.789 --> 00:22:57.109 The SEC charged a former Akamai executive with 00:22:57.109 --> 00:22:59.769 illegally tipping non -public financial information 00:22:59.769 --> 00:23:01.529 about the company. And this was part of that 00:23:01.529 --> 00:23:06.069 huge infamous scheme run by Raj Ratnam? The galleon 00:23:06.069 --> 00:23:08.450 management hedge fund scheme, yes. It's just 00:23:08.450 --> 00:23:10.970 a stark reminder that even a company founded 00:23:10.970 --> 00:23:13.279 on these pure academic math... medical principles 00:23:13.279 --> 00:23:15.839 is still vulnerable to the same human ethical 00:23:15.839 --> 00:23:17.700 failures you see on Wall Street. And maybe the 00:23:17.700 --> 00:23:19.759 most complex issue of all for a company like 00:23:19.759 --> 00:23:22.099 this is surveillance and government access to 00:23:22.099 --> 00:23:24.759 data. It is, especially given their unique architecture. 00:23:25.000 --> 00:23:27.299 And this brings us to a 2014 report concerning 00:23:27.299 --> 00:23:30.119 the NSA and the FBI. What did the report allege? 00:23:30.519 --> 00:23:32.380 It alleged that the National Security Agency 00:23:32.380 --> 00:23:35.059 and the Federal Bureau of Investigation had used 00:23:35.059 --> 00:23:38.680 Facebook's Akamai -powered CDN to collect information 00:23:38.680 --> 00:23:41.019 on Facebook users. And the implication there 00:23:41.019 --> 00:23:43.200 is that because Akamai's network is so vast, 00:23:43.440 --> 00:23:46.319 so distributed, it becomes a prime target. It 00:23:46.319 --> 00:23:48.380 becomes a prime target or a potential vector 00:23:48.380 --> 00:23:53.339 for intelligence gathering. 365 ,000 servers 00:23:53.339 --> 00:23:55.759 handling a staggering amount of the world's data, 00:23:56.000 --> 00:23:59.440 you are by definition an entity of interest to 00:23:59.440 --> 00:24:02.140 state actors. So whether the company is directly 00:24:02.140 --> 00:24:04.819 complicit or not, its very architecture creates 00:24:04.819 --> 00:24:07.519 a point of interest or a potential vulnerability. 00:24:07.940 --> 00:24:09.980 That is the essential contest here. Akamai built 00:24:09.980 --> 00:24:12.700 this network for speed and efficiency, but its 00:24:12.700 --> 00:24:15.519 success means it is now a non -negotiable piece 00:24:15.519 --> 00:24:18.079 of global communications infrastructure. And 00:24:18.079 --> 00:24:20.579 that forces it to navigate these incredibly complex 00:24:20.579 --> 00:24:23.319 legal and privacy issues that are a world away 00:24:23.319 --> 00:24:25.859 from the MIT lab where it all started. Let's 00:24:25.859 --> 00:24:27.579 circle back to that lab for a second, because 00:24:27.579 --> 00:24:29.460 despite all this corporate growth, all the controversy, 00:24:29.660 --> 00:24:31.859 their core identity is still deeply rooted in 00:24:31.859 --> 00:24:34.680 computer science. That academic foundation is 00:24:34.680 --> 00:24:37.640 still their strategic backbone. Their researchers, 00:24:37.759 --> 00:24:39.940 their founders, they've continued to publish 00:24:39.940 --> 00:24:42.920 key scientific papers that literally form the 00:24:42.920 --> 00:24:45.299 basis of how modern content delivery works. So 00:24:45.299 --> 00:24:47.420 they're still pushing the science forward. Constantly. 00:24:47.460 --> 00:24:50.079 You can find their papers on consistent hashing 00:24:50.079 --> 00:24:52.359 from the 90s, on globally distributed content 00:24:52.359 --> 00:24:54.839 delivery, and more recent sophisticated work 00:24:54.839 --> 00:24:58.400 on how their DNS system provides these authoritative, 00:24:58.440 --> 00:25:02.240 optimized answers to the world's queries. This 00:25:02.240 --> 00:25:04.480 ensures their edge is based on patented algorithms, 00:25:04.779 --> 00:25:07.440 not just on having a lot of servers. So when 00:25:07.440 --> 00:25:09.640 you put it all together, Tom Leighton, the mathematician, 00:25:09.859 --> 00:25:12.519 returning as CEO, the reestablished CPO role, 00:25:12.619 --> 00:25:14.640 the massive spending on security and compute, 00:25:14.960 --> 00:25:17.599 what does it all mean for Akamai's position today? 00:25:17.700 --> 00:25:20.380 Where do they fit in the modern Internet landscape? 00:25:20.799 --> 00:25:23.039 I think they've successfully navigated a really 00:25:23.039 --> 00:25:25.599 difficult transition. They are still, without 00:25:25.599 --> 00:25:27.700 a doubt, one of the original and most dominant 00:25:27.700 --> 00:25:30.500 content delivery networks in the world. But they 00:25:30.500 --> 00:25:32.500 are so much more than that now. They've transcended 00:25:32.500 --> 00:25:35.680 that label. Completely. They are now an established, 00:25:35.900 --> 00:25:39.160 integrated player in the very competitive cloud 00:25:39.160 --> 00:25:42.500 computing space. They sit alongside giants like 00:25:42.500 --> 00:25:44.880 Microsoft and Google, but they offer something 00:25:44.880 --> 00:25:47.940 fundamentally different. A decentralized architecture. 00:25:48.460 --> 00:25:51.279 They've morphed from just an optimization tool 00:25:51.279 --> 00:25:53.839 into a full -blown enterprise platform that gives 00:25:53.839 --> 00:25:56.940 you speed, security, and computing all bundled 00:25:56.940 --> 00:25:59.359 together. They're trying to be the entire decentralized 00:25:59.359 --> 00:26:02.279 stack. The network, the shield, and the server 00:26:02.279 --> 00:26:04.599 all operating from the edge. That's it, exactly. 00:26:04.759 --> 00:26:06.460 They're changing their value proposition from 00:26:06.460 --> 00:26:09.500 just optimizing web traffic to actually hosting, 00:26:09.799 --> 00:26:12.099 defending, and running the applications themselves 00:26:12.099 --> 00:26:15.119 where you, the user, need them. They let you 00:26:15.119 --> 00:26:17.549 bypass the late. of those big central cloud regions 00:26:17.549 --> 00:26:20.190 entirely. Hashtag tag tag outro. So let's try 00:26:20.190 --> 00:26:22.109 to synthesize this. What's the core takeaway 00:26:22.109 --> 00:26:24.349 for you, the learner, when it comes to Akamai 00:26:24.349 --> 00:26:26.529 technologies? Well, the first thing is to forget 00:26:26.529 --> 00:26:29.190 the old definition of them as just a speed booster. 00:26:29.410 --> 00:26:32.309 Right. Akamai has evolved into this fully integrated 00:26:32.309 --> 00:26:35.549 security and cloud computing platform. It acts 00:26:35.549 --> 00:26:38.250 as the essential, massively distributed traffic 00:26:38.250 --> 00:26:40.869 controller for the entire modern web. They sit 00:26:40.869 --> 00:26:43.230 in this incredibly strategic position between 00:26:43.230 --> 00:26:46.160 you and the data you need. And they're simultaneously 00:26:46.160 --> 00:26:49.200 optimizing the route, scrubbing out the malicious 00:26:49.200 --> 00:26:52.039 traffic, and in many cases providing the actual 00:26:52.039 --> 00:26:54.039 computing power that runs the service you're 00:26:54.039 --> 00:26:56.000 using. That initial mathematical breakthrough, 00:26:56.160 --> 00:26:58.420 the idea that led to this enormous network of 00:26:58.420 --> 00:27:02.420 365 ,000 servers, is now the engine that's driving 00:27:02.420 --> 00:27:05.299 this triple threat of speed, defense, and compute. 00:27:05.789 --> 00:27:08.150 And all those acquisitions we talked about, Linode 00:27:08.150 --> 00:27:10.609 for cloud, Prolexic for defense, Guardacore for 00:27:10.609 --> 00:27:13.250 zero trust security. They've all been pieces 00:27:13.250 --> 00:27:16.029 of a puzzle to secure their position as this 00:27:16.029 --> 00:27:19.150 systemic global force in the digital economy. 00:27:19.410 --> 00:27:21.849 Which brings us to our final provocative thought 00:27:21.849 --> 00:27:24.710 for you to consider. Akamai is absolutely central 00:27:24.710 --> 00:27:27.069 to how the web performs. It provides crucial 00:27:27.069 --> 00:27:30.549 speed, crucial security, but its sheer scale 00:27:30.549 --> 00:27:33.089 and the centralized control over that massively 00:27:33.089 --> 00:27:36.430 distributed system. It raises a fundamental question. 00:27:36.589 --> 00:27:38.349 A question about the future architecture of the 00:27:38.349 --> 00:27:41.170 web itself. Exactly. If the edge is everything, 00:27:41.410 --> 00:27:44.210 and one company provides both the highway and 00:27:44.210 --> 00:27:46.750 the security checkpoint for so much of the world's 00:27:46.750 --> 00:27:50.019 digital traffic, What are the long -term implications? 00:27:50.579 --> 00:27:53.619 What does it mean for the resiliency, the diversity, 00:27:53.740 --> 00:27:56.079 and maybe even the freedom of the internet when 00:27:56.079 --> 00:27:58.839 such a significant amount of control rests with 00:27:58.839 --> 00:28:02.339 one single intelligent platform? It's a powerful 00:28:02.339 --> 00:28:04.839 concentration of digital infrastructure. And 00:28:04.839 --> 00:28:06.319 I think we're only just beginning to understand 00:28:06.319 --> 00:28:08.859 the consequences of that. Something worth mulling 00:28:08.859 --> 00:28:10.819 over the next time that massive download finishes 00:28:10.819 --> 00:28:11.400 in a flash.