Topic
100 facts from 32 episodes, in date order. These are what speakers said on the shows, each linked to the moment it was said. They aren't independently verified.
1997
Amazon went public in 1997 at roughly a billion-dollar valuation, and nobody could have foreseen that it would later build AWS.
“amazon in 97 when they went public you know i think around a billion dollar valuation nobody could have foreseen aws”
2002
In 2002 Amazon held a developer conference attended by eight people and announced a new division called Amazon Web Services.
“They hold a conference for developers in 2002. A total of eight people attend the conference.”
2002
Colin Bryar, who ran the Amazon Associates program, became the first head of Amazon Web Services in 2002.
“Colin technically becomes the first head of AWS.”
2003
Andy Jassy's AWS proposal document asked Amazon to hire 57 new people to pursue the cloud infrastructure initiative.
“In the document, there is an ask, a proposal to hire 57 new people to go pursue this initiative.”
2003
Andy Jassy said a tiger team of Amazon's ten best technical minds deconstructed major web applications to decide which cloud primitives AWS needed.
“they got a tiger team together of the 10 best technical minds inside the company.”
2005
Matt Garman did a business school internship at Amazon in 2005 and joined Andy Jassy's new technology business there, which was AWS before launch.
“I did my business school internship at Amazon, and, uh, as we were looking around for projects, uh, I had talked actually at the time to Andy Jassy”
2006
Amazon S3, the Simple Storage Service, was the first AWS infrastructure service to launch, in March 2006.
“S3 is the first service to launch by itself in March, 2006”
2006
Amazon Web Services launched in 2006, which Adam Selipsky describes as pioneering cloud computing.
“when AWS, uh, launched in 2006 and pioneered the cloud, all of a sudden, there was this, uh, illusion of infinite capacity.”
2006
Matt Garman said that in 2006 AWS had to convince customers why a bookseller should run their computers.
“2006, we had to answer the question of like, why would a bookseller ever run my computers?”
2006
Matt Garman said AWS had no external customers when it started in 2006 and now has a million or more.
“When we started in 2006, you know, we had 0 external customers, and we now have a million external customers, or or multiple millions of external customers, right?”
2007
Amazon began the AWS Startup Challenge in September 2007, and Justin.tv, which later became Twitch, was a contestant in the first one.
“In fall of 2007, they start the AWS Startup Challenge and they host it first in September 2007.”
2008
Jeff Bezos spoke at Y Combinator's 2008 Startup School shortly after Amazon Web Services launched, pitching startups on using AWS.
“in 2008, Bezos came and this was right after AWS had launched”
2008
AWS launched the CloudFront content delivery network in 2008 and its first major database offering, the Relational Database Service (RDS), in 2009.
“CloudFront, which is the content delivery network, the CDN, the Akamai part of the puzzle piece, that launches in 2008. And the first major database offering, RDS, the relational database service, doesn't actually launch until 2009.”
2008
Around 2008 or 2009 AWS staff visited financial firms in New York, including Goldman Sachs and JPMorgan, who were fact-finding about cloud computing.
“We went on a trip to New York, and actually, there's a bunch of financial services com companies, the Goldman Sachs' and JPMorgan's”
2009
Netflix became an AWS customer in 2009.
“2009, Netflix becomes a customer. And how crazy is this?”
2010
A host says AWS revenue grew from about $500 million in 2010 to about $90 billion in the most recent year.
“I think you guys went from something like 500 million or so in 2010 to about 90 billion last year in terms of revenue for AWS.”
2011
Werner Vogels wrote in a 2011 Quora post that Amazon Web Services was never built by selling excess capacity and that AWS would have exhausted Amazon.com's spare capacity within two months of launch.
“It was never a matter of selling excess capacity. Actually, within two months after launch, AWS would have already burned through the excess Amazon.com capacity.”
2014
Jeff Bezos wrote in his 2014 shareholder letter that he believed AWS's market size was unconstrained.
“I believe that AWS is market size unconstrained.”
2015
Amazon revealed financial results for Amazon Web Services in 2015, after which its profitability became widely recognized.
“AWS was a cipher in 2013. The company revealed the financials in 2015. And the world recognized how good of a business that was.”
2015
When Amazon first broke out AWS financials in 2015, AWS had about a $6 billion revenue run rate and 70 percent annual growth.
“IPO, quote unquote, $6 billion revenue run rate for AWS, 70% annual growth rate.”
2015
AWS made 23 price reductions by 2012, 40 by 2013 and 51 by 2015, without strong competitive pressure.
“They had already done 23 price reductions across the board for all of their services. By 2013, they had done 40. By 2015, they had done 51.”
2019
Amazon.com did not finish migrating off Oracle databases onto AWS database products until 2019.
“Amazon.com did not finish their migration off of Oracle databases and onto AWS products until 2019”
2020
Amazon suspended police use of its Rekognition image-recognition service for a year amid concerns about bias, and Brad Stone believes it may now be suspended indefinitely.
“They suspended it for a year after the sort of outcry and anxiety about image recognition.”
2020
Supabase spent the first week of its Y Combinator batch migrating about 800 databases from DigitalOcean to AWS.
“And so we spent the first week of being in YC, migrating 800 databases to a different cloud, which, you know, if you've ever migrated one database, you know how painful it is.”
2021
Andy Jassy succeeded founder Jeff Bezos as Amazon CEO in 2021 after running Amazon Web Services.
“He succeeded Jeff Bezos, who was the founder CEO of Amazon. He ran Amazon Web Services, which grew phenomenally. You know, he's been the CEO since 2021.”
2022
David Rosenthal said AWS was growing 37% on a $75 billion base, while Google Cloud and Microsoft Azure grew about 45% but had a combined share smaller than Amazon's.
“AWS is growing at 37% annually on a $75 billion base. Google and Microsoft are growing at 45% but their market share combined is still significantly less than Amazon.”
2022
The host says AWS's reported revenue backlog of contracted but not yet recognized revenue exceeded $100 billion.
“That backlog of committed contractual signed revenue is over $100 billion.”
2023
Adam Selipsky estimated that roughly 10% of global IT spending had moved to the cloud, and later put it at 10-15%.
“I'd estimate probably somewhere in the realm of 10% of IT has moved to the cloud”
2023
Adam Selipsky said over 80% of the world's unicorns, startups worth over $1 billion, run on Amazon Web Services.
“over 80% of the world's unicorns, uh, run on AWS, and it's a great, great privilege to work with all of those innovators.”
2023
Amazon Web Services operated 32 independent infrastructure regions worldwide.
“So we have 32 independent infrastructure regions around the world, uh, many in the US and Canada”
2023
Each Amazon Web Services region has at least three availability zones, which are built to fail independently of each other.
“E- each one has at least 3 availability zones per region, and that's a really, really important set of concepts”
2023
Adam Selipsky said Amazon Web Services was over a $90 billion a year revenue business.
“we're we're over a 90 billion dollar a year revenue organization”
2023
Amazon Web Services says it has lowered prices more than 120 times in the 17 years since launch.
“We've lowered prices over 120 times, uh, in the past 17 years, uh, since we launched.”
2023
Anthropic uses Amazon Web Services as its primary cloud provider, and Amazon has invested in the company.
“Anthropic, um, now uses AWS as its, uh, primary, uh, cloud provider”
2023
Tom Davenport said a survey he ran with MIT CDO IQ and AWS found that about 80 (apparently percent) expected generative AI to transform their organizations, but only six had a production application.
“80 said they thought it was going to transform their organizations, but only six had a production application that they had deployed.”
2024
Garman says about 85% of computing workloads still run on premises, by most estimates.
“I think 85% of workloads are still running on prem today by most estimations, somewhere in that range”
2024
AWS made Meta's Llama 3.1 available on Bedrock on the day it launched, and Garman describes a tight partnership with Meta.
“we just launched, uh, Llama 3.1, uh, on the day it launched, and we have a really tight partnership with Meta”
2024
Garman says Adobe built its Firefly models on top of AWS.
“we see folks like Adobe building Firefly as a all built on top of AWS, purpose built for their own thing”
2024
Matt Garman said Amazon Web Services had grown into a $110 billion business growing at 19%.
“I think over eighteen years, we've grown to be 110 billion dollar business, growing at 19%.”
2024
After the interview was recorded, Amazon announced an $11 billion investment to expand its data center operations in Georgia.
“after Matt and I had this conversation, the company announced 11 billion dollar investment to expand its data center operations in Georgia.”
2024
Matt Garman said Netflix runs its own homegrown content delivery network and that this part, not its AWS infrastructure, was stressed during the Jake Paul-Mike Tyson stream.
“Netflix has their own homegrown CDN that they use, and that's the part that I think was stressed.”
2024
At its re:Invent conference, AWS announced Q Business Index, which pulls a company's SaaS data into a single index that the enterprise owns and controls.
“just at re:Invent, we announced a capability called Q Business Index, where you can basically have all of your SaaS data pulled together in a single index that's owned by the enterprise and controlled by the enterprise”
2024
Matt Garman said many of AWS's largest customers have only 10% to 40% of their workloads in the cloud.
“You can even take our largest customers, many of them only have ten, twenty, thirty, 40% of their their workloads in the cloud.”
2024
Matt Garman said AWS has about 200 different services.
“There's 200 different services, so I'm not gonna go through all of them”
2024
Matt Garman said the newest generation of AI hardware, including AWS's Trainium2 and Nvidia's Blackwell GB200s, should give customers a two to four times boost in compute per dollar.
“we just announced with Trainium two, and eventually, uh, with, uh, Blackwell and and GB200s, I think will give customers a two to four x boost in compute capability per dollar.”
2025
Hershey says he has seen Sonnet 4.5 deploy a complex application to AWS autonomously and run a security audit on it.
“I've seen it be able to deploy a complex application to AWS, like fully autonomously, and do like a security audit on it.”
2025
Chandrasekar attended AWS re:Invent in the week of the interview, where Amazon discussed its Trainium chips, including Trainium 3 and Trainium 4.
“I was at, you know, AWS reInvent this week where they talked about the Trainium chips and Trainium three and Trainium four being built out.”
2026
Tim Crawford names IBM's Granite, Amazon Web Services' Nova and Google's Gemini as models whose makers are building in safety measures.
“whether it's granite from IBM, Nova from AWS, Gemini from, uh, Google, for example, they are putting safety measures in the mix.”
2030
Matt Garman said AWS is excited about small modular nuclear reactors, which he sees as a technology still a little ways away that could matter after 2030.
“we're excited about, uh, small modular reactors. I think that that it's a technology that's a little ways away”
Undated
Brad Stone says Amazon's retail business is the biggest and first customer of Amazon Web Services.
“Retail is the biggest customer of AWS. It's the first customer.”
Undated
Amazon targeted gross and operating margins of 20 to 40 percent for AWS, which the host says was unattractive to incumbents earning around 80 percent gross margins.
“Amazon targeted gross margins and operating margins for AWS in the 20 to 40% range.”
Undated
AWS created the Snowball, a 100-terabyte secure device shipped to customers' offices, to move large datasets into its cloud.
“they first released this thing that was a hundred terabyte, super secure thing they would ship to your office called the Snowball.”
Undated
The host says AWS earns gross margins above 90 percent on bandwidth charges.
“AWS makes 90 plus percent gross margins on their bandwidth charges.”
Undated
One host says AWS veterans told them that AWS does not deprecate features, unlike other Amazon products.
“that became very clear that AWS doesn't deprecate features Amazon will kill stuff”
Undated
Chenyi Shi noted that Amazon, despite AWS's success, flopped with the Fire Phone and lost billions of dollars on Alexa.
“But as successful as they are in AWS, they also flopped Fire Phone, if you still remember that. They also lost billions and billions on Alexa.”
Undated
Davenport conducted a second annual survey with the MIT Chief Data Officer Group Symposium, sponsored by Amazon Web Services.
“I was just looking at a survey that I did with the, um, MIT Chief Data Officer Group Symposium, the second annual survey sponsored by Amazon Web Services and data just came in.”
Undated
Early working-backwards documents for Amazon Web Services stated the goal of giving a student in a dorm room the same computing infrastructure as an Amazon engineer.
“we wanted the student in a dorm room to have access to the same world class computing infrastructure as any Amazon software engineer.”
Undated
Jeff Bezos publicly called AWS's 5-7 year head start over competitors the largest gift in business history, according to Adam Selipsky.
“our founder, Jeff Bezos, has actually publicly called that 5 to 7 year head start the largest gift in business history.”
Undated
Adam Selipsky cited AWS's role in controlling NASA's Mars Rover as evidence that mission-critical workloads run in the cloud.
“controlling, uh, NASA's Mars Rover, you know, SUV-sized vehicles on the surface of Mars”
Undated
Adam Selipsky said AWS's in-house Graviton chip uses 60% less power than equivalent chips.
“Our our Graviton-based, uh, um, uh, chip, um, actually consumes 60% less power than equivalent, uh, equivalent chips that our customers use.”
Undated
Adam Selipsky said AWS was founded partly because Amazon wanted its teams to move faster by sharing a common layer of infrastructure services.
“one of the reasons why AWS was founded was because Amazon wanted to move faster, and they wanted to create the shared layer of infrastructure services”
Undated
Jeff Bezos says Amazon Web Services is building a foundation model called Titan and a corporate tool called Bedrock.
“we're building Titan, which is our, uh, foundational model, we're also building, y- um, Bedrock”
Undated
Matt Garman returned to Amazon full time as effectively the first product manager for AWS and says he has been there eighteen years.
“I came back full time, uh, effectively as the first product manager for AWS, and I'm now on year eighteen”
Undated
Garman says an early AWS virtual Linux server arrived in thirty seconds, when getting a physical server normally took about six months.
“it auto scaled and you got it in thirty seconds, which was pretty awesome at the time, where normally it was gonna be six months or something for you to go get a server”
Undated
Garman says AWS deliberately targeted the hardest workloads, such as JPMorgan Chase and US intelligence agencies, so that regular companies' objections would shrink.
“if I can solve the, you know, JPMorgan Chase running in AWS, if I can run solve the US. Intelligence Agency running in AWS, like the the reasons for regular other companies are diminishingly small.”
Undated
About a year into working on AWS, Garman told a friend he thought it could become a billion-dollar business.
“Oh, how's that AWS thing going? And I was like, You know what? I think this thing could be a billion dollar business.”
Undated
AWS won a secret cloud contract with US intelligence agencies, competing against incumbents such as HP, IBM and Oracle.
“we went after the the intelligence agencies for the US. government, and we won that contract and it was secret”
Undated
Garman says IBM sued over the AWS intelligence contract, which made the win public, and the agencies then publicly called AWS the most technically sophisticated option.
“so then it became public that we won this deal. And then the intelligence agencies went out to public and said, no, AWS is the most technically sophisticated”
Undated
Garman says AWS has invested in AI for about ten years and began building its own processors five or six years ago.
“we've been investing in AI broadly for the last ten years, and that's why we've started five or six years ago investing at the infrastructure layer and building our own processors”
Undated
Garman says AWS's AI strategy assumed there would be many models of different sizes and purposes, which customers would combine.
“We thought that there was gonna be a lot of models for a lot of different purposes, and there'd be big models and small models, and people would want to combine them in new and interesting ways.”
Undated
AWS partners with Scale AI to help customers label data for fine tuning or pre-training.
“we have a deep partnership with Scale AI to come in and help you label your data”
Undated
Garman says AWS has been the largest purchaser of new renewable energy contracts in each of the last four or five years.
“we're the the largest purchaser of of renewable energy, um, over the last, uh, new new contracts, right?”
Undated
Garman says NVIDIA runs its own AI training clusters on AWS because AWS has the most stable infrastructure.
“NVIDIA actually runs their AI training clusters in AWS because we actually have the the most stable infrastructure of anyone else.”
Undated
Garman says AI workloads are roughly half training and half inference today, but inference must dominate for investment in big models to pay off.
“in order for the math to work out, inference workloads have to dominate. Otherwise, all this investment in in these big models isn't really gonna pay off.”
Undated
Garman says some of the largest AI models are reported to need gigawatts of data center capacity.
“they're talking about needing gigawatts of capacity, um, which is, uh, is is kind of a a mind boggling number”
Undated
Garman says building capacity for the largest AI models involves tens, if not hundreds, of billions of dollars of capital investment.
“we're talking tens, if not hundreds of billions of dollars of of capital investment”
Undated
Bret Taylor says cloud computing consolidated into a small handful of infrastructure-as-a-service providers, such as Amazon Web Services, Azure and Google Cloud, that account for most cloud capital spending.
“you ended up with a small handful of infrastructure as a service providers that represent the vast majority of the capex investment in cloud”
Undated
MacInnis says Rippling has built a large-scale data platform that it compares to AWS and expects to launch more around it soon.
“it's kinda like our AWS, like, we're gonna have our AWS moment at some point in the next, you know, quarter or so”
Undated
Matt Garman said AWS's Verizon arrangement was a partnership to potentially sell edge compute space together, rather than Verizon building a rival cloud.
“Well, Verizon was a little bit different. It was actually a partnership with us where we were talking about potentially selling some of that compute space together at the edge.”
Undated
Matt Garman said AWS organizes product work into teams of ten to twenty people, each owning a fully separable subcomponent of a service.
“how do we have teams of of ten to twenty people that are are focused on a subcomponent of that that that is is fully separable.”
Undated
Matt Garman said AWS has been making its own custom silicon chips for more than a decade.
“we've been making chips now for over a decade. Um, so we're we've been making silicon chips, our own custom silicon, for more than a decade.”
Undated
Matt Garman said Graviton showed AWS can launch its own processor at broad scale, while AWS still buys many Intel and AMD chips.
“We've proven that we can launch a processor at broad scale that is very useful for a set of workloads”
Undated
Matt Garman said AWS partners with Samsung, Intel and other companies that have their own chip fabrication plants.
“we partner with Samsung, we partner with Intel, we partner with others that have their own fabs as well”
Undated
Matt Garman said power transmission, not just generation, is a bottleneck for AWS data centers.
“Another of the bottlenecks that we run into is is around transmission. It's not just power generation, but it's transmission.”
Undated
Matt Garman rejected the idea that AWS began as a way to resell Amazon's spare capacity, saying the math does not work.
“AWS was never about excess capacity about about Amazon. It just just, like, math doesn't work, right?”
Undated
Marc Benioff said Salesforce has been rebuilt to run on cloud infrastructure from Amazon, Alibaba and Google.
“We've been rebuilding Salesforce so that we run well, for example, in Amazon and Alibaba and Google”
Undated
Linda Yao says that about a decade before the episode, Amazon Web Services made computing more accessible, enabling use of big data.
“Then we fast forward ten years to about a decade ago, and that was around the time when AWS really made computing more accessible and democratized.”
Undated
Andy Jassy says Amazon's EC2 compute service started with about eleven people.
“We had um eleven people when we or thirteen people when we started our compute service EC2. We had eleven people.”
Undated
Andy Jassy says Amazon's SageMaker service has become a standard way for people to build their own AI models.
“we've built a service called SageMaker, which is the really kind of become the standard way for people building their own models to get the data in”
Undated
Andy Jassy says Amazon's AI services contribute to a multibillion-dollar annual revenue run rate in AI.
“these are services that are part of why we have a multi billion dollar annual revenue run rate in the AI space.”
Undated
On one AWS-based engagement, prompts were submitted by dropping a file into an S3 bucket that triggered a Lambda, with results appearing in another S3 bucket.
“And the way you introduce the prompt is by dropping a file on an S3 bucket, calling a Lambda, and then in another S3 bucket.”
Undated
Before Stack Overflow, Chandrasekar worked at Rackspace in cloud services, where he ran a business responding to the threat from Amazon Web Services.
“my experience right before that was at Rackspace, uh, in the cloud services space.”
Undated
Anima Anandkumar was previously principal scientist at Amazon Web Services and senior director of AI research at NVIDIA.
“Previously, she was principal scientist at Amazon Web Services and senior director of AI research at NVIDIA.”
Undated
Cannon-Brookes says AWS's Matt Garman told him at an event that AWS would also have more developers in five years.
“I sat on a, uh, I I spoke with with Matt Garmin at AWS at an event, and he said exactly the same thing.”
Undated
CJ Desai says the cloud transition that began with AWS is approaching its twentieth year and is still under way.
“if cloud started with A W S, we are in the almost approaching twentieth year, and it's still going on.”
Undated
AWS Bedrock can run a DeepSeek model on its own servers as a managed service.
“This would be like AWS Bedrock runs a DeepSeek model in their servers.”
Undated
McLuckie gives the example of configuring the AWS MCP server in read-only mode by exchanging an Okta token for a scoped AWS token, to prevent an agent from deleting an RDS instance.
“I want to use the AWS MCP server and I want to use in read only mode, right?”
Undated
In Datadog's sandbox approach, coding agents cannot read credential files such as AWS or NPM publish secrets in a developer's home directory.
“So even if you have a file in your home directory that has your AWS secret, your NPM publish secret or what have you, the agent on the sandbox can't touch it.”
Undated
MongoDB's managed service, Atlas, can be deployed in the data centers of Amazon, Google and Azure.
“which you can deploy your instances on any cloud hyperscaler data center you'd like amongst Amazon, Google, and Azure”
Undated
Das Deb says connected applications often run on different technology stacks, such as Java, Python and C, as well as AWS and AI components.
“One is running on Java, one is running on Python, one is running on C, and then we have A W S coming in, then we have A I coming in.”
Episodes