Cast AI Data Shows Kubernetes Users Are Spending Too Much Money

Hello and welcome to Protocol Enterprise! Today: why most Kubernetes users spend far more on compute resources than they need, Google collects all data from Sidewalk Labs and why HashiCorp CEO Dave McJannet thinks you need a team cloud platform.

Hit by the clusters

According to Cast AI, cloud customers pay on average three times more on cloud compute costs for AWS, Microsoft Azure and Google Cloud than they should. Helping them manage these costs becomes a business in itself.

The startup specializes in Kubernetes automation, cost optimization, and reporting for cloud-native applications. Its platform uses artificial intelligence to identify the compute resources needed for specific Kubernetes workloads and automatically selects the best combinations, configuring processors and memory to avoid overprovisioning.

It continuously adds or removes resources as needed, ensuring customers don’t overspend without compromising workload availability or performance, according to the company.

  • “It’s impossible to do this exercise as a human,” said co-founder and product manager Laurent Gil. “We decomplex capacities. We are making Kubernetes or containers serverless by saying we will take care of the servers and we will make the servers profitable. »
  • Cast AI was born out of its co-founders’ frustrations with their cloud bills while operating a previous startup: Zenedge, a cloud-based, AI-powered cybersecurity startup acquired by Oracle in 2018.
  • “When I started this business, I was spending about $1,000 to $2,000 a month on AWS,” Gil said. “Three years later… it became $2 million – by far the highest cost in the business, and we were very, very frustrated. We had a good journey with customers, but every time we added a customer, our AWS bill exploded.
  • AWS’ response was for Zenedge to prepay for three years to reduce its cloud bill by 40%, but Zenedge didn’t want to be locked in, according to Gil. With Cast AI, he and his co-founders built the expense management product they wanted to have at the time.

Businesses using Cast AI services can reduce their cloud compute spend by an average of 65%, according to Gil. These services work with Amazon Elastic Kubernetes Service, Google Kubernetes Engine, Azure Kubernetes Service, and Kubernetes Operations on AWS.

  • “The engine will instantly understand what apps you have…and how much compute and memory they are currently consuming, and how much they cost to run based on the machine those apps are installed on,” Gil told Protocol.
  • While there aren’t big price differences between the big three cloud providers, Gil said that within each cloud itself, there are cost differences when it comes to processors.
  • “Most … are cheaper with AMD than with Intel,” Gil said. “This causes our engine to sometimes use more AMD for computationally intensive [workloads]. But the machine was trained to know that, so we’ll always choose the cheaper option.
  • “One thing that really surprised us…is that the average savings we offer anyone using us…is 65%,” he said.

Read the full report here.

— Donna Goodison (E-mail | Twitter)


Accounts Receivable is essential to a positive customer experience, but many leaders ignore it. This limits the potential for success and gives competitors the opportunity to get ahead. Discover how collaborative AR not only optimizes your results, but also gives you a unique advantage in the market.

Learn more

Google sucks data from Sidewalk Labs (and Doctoroff returns to building the future)

When Alphabet’s “city of the future” Toronto sidewalk project imploded in 2020, it was a thud heard around the world. But several data-hungry urban tech companies have crawled under the wreckage and survive today. Sidewalk subsidiaries, including “smart” parking company Pebble, energy technology company Mesa and automated timber construction provider Canopy Buildings, joined Google late last year.

Now the data transfer has started.

“We are writing to inform you that, effective June 30, 2022, your information will be transferred from Sidewalk Labs to Google LLC,” said an email sent to people (including this reporter) who had signed up for the updates from Coord, the Sidewalk company now named Pebble.

Expect the data transfer to include more than just names and email addresses, though. Like many Sidewalk offshoots, Pebble wants to digitize urban environments. This means new datasets associated with urban spaces – What’s on that sidewalk? – to use these spaces more efficiently or help delivery companies solve this “last mile” problem.

Oh, and a “parallel” note: when Sidewalk founder and CEO Dan Doctoroff – former New York Deputy Mayor for Economic Development under Mayor Mike Bloomberg and former CEO of Bloomberg LP – announced the reorganization of Sidewalk in December, he announced he would be leaving his role with the company, revealing that he may have ALS.

At the time, he said he intended to spend time babysitting his grandchild and “finally learn[ing] speak French fluently. Today, a few months later, he is starting to build the future. Doctoroff has joined a new blue ribbon panel set up by New York Mayor Eric Adams and New York Governor Kathy Hochul to “develop concrete strategies for the resurgence and resilience of the city’s commercial districts”.

-Kate Kaye (E-mail | Twitter)

HashiCorp: You Probably Need a Cloud Platform Team

As COVID-19 constraints eased and allowed for more travel in the last quarter, HashiCorp CEO Dave McJannet and co-founder and CTO Armon Dadgar spent much of the period in visit the company’s Global 2000 customers in person. According to McJannet, they heard a recurring theme: the emergence of centralized cloud program offices or cloud platform teams.

While such platform teams — sometimes called cloud centers of excellence — are prevalent in cloud-native enterprises, the concept has yet to take off with the majority of the Global 2000, according to McJannet.

“We have this weird perch in the middle of the cloud market as a kind of enabling technology that everybody uses to interface with clouds, so we kind of sit in the data stream of what the approaches look like everyone’s cloud,” McJannet told Protocol. “The technical issue of how you run the cloud is actually pretty well understood, but there’s actually an organizational shift that we’re seeing in all of these companies as well.”

McJannet described it in terms of a “cloud 1.0” versus “cloud 2.0” approach.

“In the cloud 1.0 approach, organizationally, I’d say, ‘Hey, go build me an app on Amazon,’ and your developers are going to build an app on Amazon,” McJannet said. “You think you’re happy, but 12 months later you realize your Amazon bill is way overdue, and there’s a bunch of apps and the security team is like, ‘What the hell is this? What are you doing? It’s not a private network. This all works on the public internet. We’re going to end up on CNN – stop.’ »

This is where cloud platform teams come in to standardize how developers interact with the cloud and provide greater control in accordance with policy and governance.

“I would say for anyone who is a successful cloud adopter, yes, it’s widespread,” McJannet said. “But I would say most people are still very early in their journey to cloud. All cloud-native companies, like Slack or Stripe, are all built that way. They have a platform team, and then they have developers. But I would say in the Global 2000, no, it’s probably only about a third or half of them that have somehow progressed to that level of maturity.

—Donna Goodison

Around the company

ServiceNow acquired Hitch Works, a company that promises to help management find out which of their own employees have the skills needed for major projects, for an undisclosed amount.

Cloud infrastructure services grew 41.4% in the first quarter to $90.9 billion, according to Gartner. AWS remains well in control of the market, but Google Cloud is growing the fastest.


It’s hard to find a leader who would say the customer experience isn’t a priority for them. Yet only 44% see better communication with customers as a benefit of augmented reality digitization. This represents a huge opportunity, and collaborative AR is the key to seizing it.

Learn more

Thanks for reading – see you Monday!