From the Archives - A Decade of Working Backwards
I have now been at Amazon Web Services for fifteen years. It is a journey I never expected and one of the most fulfilling I could have experienced. Just a month ago, I stepped away from the org I started nine years ago to start a brand new journey leading our generative AI-driven builder experience. I am not going to spend too much time reminiscing, but wanted to repost something that was originally written on Medium at my ten year mark. It still applies
This picture was taken 6 years ago (Ed. Now 11 years ago). That particular view has changed a lot since then. While the city changes around us, I have essentially the same view from my office window (just with a lot more cranes and high rises, and less of the Needle). Just a shade over 10 years ago I walked through the doors of PacMed wondering how long I’d work at Amazon Web Services. It was my first day, and during orientation almost no one in the room knew what AWS was. I had been a fanboy pretty much from the day that S3 was announced; perhaps why someone took a chance on the the theoretical chemist whose career till that point had been in molecular modeling/drug discovery and bioinformatics. It was a job in business development for Amazon EC2. In addition to never actually working at a tech company, I barely understood what business development was (I had spent most of my career either writing algorithms or as a product manager). The very fact that I had applied for the job was a random event. I had run into Jeff Barr at a few events (notably Gnomedex) and during one late night Twitter conversation, he convinced me to send him my resume in “140 characters or less”.
Well that was 10 years ago. These are 10 years that I wouldn’t trade for anything. I get into work every morning just as excited and optimistic about the future as I did a decadeago, even though my job is very different. In fact, it is that opportunity to do different things that is one of the best parts about working here. In the past, after 2–3 years, doing the “next thing” often meant going elsewhere. Here it has always been around the corner, even though I’ve essentially stayed with EC2 my entire career at AWS. First in business development, then in product management (mostly on EC2 instances), to starting the container services org, to working on multiple product and business areas, it’s been a phenomenal ride. AWS is bigger, Seattle is bigger, and my waistline is definitely bigger. Along the way I’ve become a dad, developed interesting new hobbies, and moved houses, but some things have not changed, the things that make this the best place I could ever imagine working
That famed customer obsession. It is real. It is the mindset that allows us to build products that don’t focus on the how, but the why.
The ability to drive ideas. I’ve had the opportunity to present new ideas to leadership multiple times and, even better, the opportunity to bring some of those ideas to fruition.
Documents. I love writing and reviewing PR/FAQs, I love writing and reviewing 6-pagers. I can’t imagine working any other way.
The people. I’ve effectively worked for and with a small group of people for a decade. They’re awesome and insanely smart. From them I have learnt most of what I know about operational excellence, leadership, big thinking, and how to ignore the noise and focus on what’s important. I’ve made friendships that will last a lifetime.
My team. Their dedication, discipline, customer focus, and operational chops are humbling. More than anything else knowing that I get to work with this group of people keeps me inspired and motivated.
Our customers. Some have become friends, all push us to be better. There would be no AWS without the customers we have and trying (sometimes multiple times) to build something that solves their problems makes the most frustrating days worth the effort.
This last decade has been full of so many memorable things. I’ll list list a few of my favorites
CC1 — my first real product launch at AWS. The best part. A lot of the ideas we played around with (HVM, processor and processor topology visibility, a 10 gig network, etc) all became standard EC2 features a few years later.
CC2 — The follow up to CC1 and the time a cluster of EC2 instances entered the top 50 in the Top500.
C3 — Apparently I like my compute instances, but this one had enhanced networking.
The first re:Invent. That’s when the impact that AWS was making in the technology industry really hit me.
ECS — Who knew containers would become a thing, but this one will forever be special because of how it came to be and because it was the first time I got to build an entire business and org from the ground up.
Batch — I’ve always been the HPC and batch computing person at AWS, so it was great to get asked to go build a dedicated service for batch computing.
Nitro. I was involved a lot in the early days, but had moved on to container land by the time it shipped. It was great being able to finally talk about the amazing work that the collective team had done.
Fargate — I wrote the product definition for Fargate before I wrote the one for what became ECS. It took a few years and a lot of learning to figure out how exactly we wanted to build it and what it should look like. I am glad we waited.
EKS. It’s another great example of just listening to your customers.
There are so many more. Can’t wait for what’s next.
… experiment patiently, accept failures, plant seeds, protect saplings, and double down when you see customer delight. A customer-obsessed culture best creates the conditions where all of that can happen.
— Jeff Bezos
