A photocopier.
Photo: Influx Productions/Getty Images

Why AWS and Google Cloud copied each other

Protocol Enterprise

Welcome to Protocol | Enterprise, your comprehensive roundup of everything you need to know about the week in cloud and enterprise software. This Thursday: why cloud service portfolios are starting to look more and more alike, why Taylor Swift concert tickets might depend on edge computing, and a farewell to one of the most notorious products in Microsoft history.

(Was this email forwarded to you? Sign up here to get it in your inbox every week.)

The Big Story

I'll have what they're having

AWS and Google Cloud, like most enterprise tech companies, insist that their product-strategy decisions are driven by customers, not competitors. Yet it's not hard to tell when those decisions are driven by customers who covet a competitor's cloud service.

Both companies announced new services this week that resemble ideas refined for the mainstream by the other. AWS's App Runner, a serverless container management tool, and Google's Vertex AI, a managed machine-learning service, are clear nods to services that have been available on the other company's cloud for a few years.

Neither service comes as a shock, even if you've endured countless briefings and conferences where cloud providers barely acknowledge the existence of the other companies in the market. But it does show that they aren't afraid to copy what works, right as it becomes easier to operate a multicloud strategy.

App Runner was announced for the first time Tuesday through a press release, a signal that AWS considers the launch a major new effort.

  • App Runner is a serverless container-management tool akin to Google Cloud Run, first launched in 2019.
  • AWS has invested a great deal of time and energy in serverless container tools, but App Runner is being promoted as the easiest container management tool within a lineup of several options.
  • Some businesses want tight control over how their apps run and want the flexibility to make changes, while the target audiences for App Runner (and Cloud Run) want the benefits of containers without the hassle of managing complicated tools like Kubernetes.
  • AWS was a little later than Google (and Microsoft) to acknowledge the value of containers but has made up for that delay, and has now announced 11 container services, eight of which are available today.

Google unveiled Vertex AI this week as part of Google I/O, which historically focuses more on Google's consumer-oriented developer tools but made a little room for the enterprise this year.

  • Most companies think they want to introduce artificial intelligence or machine learning into their products and services, but quickly realize they have little to no idea what they are doing and therefore see little to no return on that investment.
  • Vertex AI was designed to make it easier to get up and running with machine-learning models that can train the datasets they already know how to collect.
  • If that sounds familiar, it's because AWS has been promoting a similar service called SageMaker for several years: Craig Wiley, the director of product management for Vertex AI, actually held a similar role at AWS during the years it launched its SageMaker tool.
  • One of Google's main advantages among rival cloud vendors to date has been its data and AI tools for experts, but as it adds more and more "normal" enterprise companies that don't have a lot of in-house AI talent to its customer list, like AWS it will need to help those less-experienced companies get started.

Most enterprise tech companies take pride in their products, but their leadership takes just as much (if not more) pride in their profits.

  • Services like Cloud Run and App Runner are high-level managed services that trade convenience for flexibility, and make it harder for customers that write applications around those services to move those apps down the road.
  • Services like Vertex AI and SageMaker are gateway drugs to the full-blown AI experience, which requires a lot of computing power and storage capacity to do properly.
  • Lots of customers are happy to pay for these conveniences and some assistance, especially the late-arriving cloud adopters that can't necessarily attract top-tier tech talent to manage these complicated tasks themselves.

Microsoft will get its turn next week during Build, its annual developer conference, to show off what it has been working on this year for Azure customers.

  • The last several years Microsoft has blasted out a firehose of incremental updates to Azure and its various enterprise software products, as you can see from the 2020 "Book of News."
  • It's usually hard to pull a theme from that cacophony (which, to be fair, also accompanies AWS re:Invent and Google Cloud Next), but with the pandemic subsiding in the U.S., expect to hear (a little) less about its pandemic response compared to last year, and more about its roadmap.
— Tom Krazit


Many believe AI is going to revolutionize health care, from clinical applications in areas such as imaging and diagnostics to workflow optimization. Confidential computing protects the privacy of patient data by enabling a specific algorithm to interact with a specifically curated data set which remains, at all times, in the control of the health care institution that developed it.

Learn more

This Week On Protocol

Cutting edge: It's going to take several years for one of Microsoft's favorite topics — edge computing — to really show its promise, according to StackPath CEO Kip Turco, but the groundwork is being laid by several companies. One simple example: a ticket-buying application built around an edge-computing service could make it much easier to actually secure a ticket to the first post-pandemic concert from your favorite big-name artist.

Start small: Part of the thinking behind services like Vertex AI and AWS SageMaker is that too many companies try to bite off more than they can chew when starting out with AI. "Learn to walk first," said Landing AI and Coursera founder Andrew Ng in an interview with Protocol's Joe Williams: "It's fine that the first project you do is not a $10 million AI project."

Creative financing: Raising traditional VC money has plenty of downsides, especially for founders worried about losing control of their companies. Pipe just raised $250 million to help companies trade future revenue streams for upfront capital, which could help enterprise software companies built around subscription services build out their services.

Five Questions For...

Mark Porter, CTO, MongoDB

What was your first tech job?

I programmed, taught and consulted all through my teenage years. My first major job was a contract with the U.S. Department of Education — I needed this to help pay for my tuition at Caltech. During my freshman and sophomore years, I wrote a course-authoring and delivery system which taught Alaskan Natives in remote villages how to be paralegals or accountants. This allowed people with no access to higher education to get vocational training and make their way in modern society. It taught me that you actually can do social good and make money at the same time.

What was the first computer that got you excited about technology?

I started programming at 11 years old on a 4K RadioShack computer and at 14 on an HP 64K calculator. The calculator program was pretty unique: Rich with oil revenues and also very cold, the Alaskan government was quite interested in people insulating their houses better and were willing to pay for it. With my program, you could walk around your house, enter all the info on your existing doors, windows and walls. The program would calculate the maximum energy savings you could achieve through upgrades, and would print out the paperwork you needed to apply for the grant as well as the information needed for a contractor to bid for the work.

What's your favorite pastime that doesn't involve a screen?

I read quite a bit and also enjoy quality time with my wife and five children. My favorite sci-fi authors are Poul Anderson and Larry Niven ("Ringworld"). I'm currently working my way through the massive eight-volume "Expanse" series. When I'm not doing that, I go kayaking at sunset on a lake near our house. Someday I plan to learn to play the piano.

Which enterprise tech legend motivates you the most?

Grace Hopper (1906-1992). For those who don't know, Grace was one of the most influential people in computer science. She was one of the first serious programmers and a true academic at both Yale and Vassar. She not only came up with great technology, such as the compiler, but she was able to do it when people didn't think compilers were either possible or useful. She helped invent COBOL, the first attempt at a computer language that humans could easily understand — and the first one that was designed to be hardware-independent.

What will be the greatest challenge for enterprise tech over the coming decade?

For decades, enterprises have maintained systems of record and systems of engagement. Systems of record are foundational, mission-critical, sources of truth that are accessed primarily by internal programs and users — they are the backbones of a business. Systems of engagement are the digital interfaces with which customers, vendors and employees interact — in applications, on mobile and on the web. For a long time, each system lived on different computers, had different data management requirements, and were funded by different departments. But that is changing.

This convergence is going to be one of the most challenging things that we've done; we need to maintain transaction rates and consistency while allowing both realtime and offline analytics. And we need to do it economically — which means we need elastic compute resources accessing elastic storage resources at every level of our infrastructure.

Around the Enterprise


Many believe AI is going to revolutionize health care, from clinical applications in areas such as imaging and diagnostics to workflow optimization. Confidential computing protects the privacy of patient data by enabling a specific algorithm to interact with a specifically curated data set which remains, at all times, in the control of the health care institution that developed it.

Learn more

Thanks for reading — see you Monday.

Recent Issues