How Cloud Puts the Engine in Platform Engineering

After graduating from college in the late aughts, Dan Kuebrich worked as a software engineer for an online startup in Queens, N.Y. Fifteen years later, he remembers what it was like for the company to manage its own IT infrastructure.

“If one of your hard drives died, you’d take the subway over to Manhattan and physically go into the data center and pull it out and put in a new one,” he says. “Looking back, it’s unimaginable — not when you have so many other important things to do.”

One reason Kuebrich feels this way today is that his current employer, Fullstory, does all of its software development in the cloud. The Atlanta-based company is the creator of a behavioral data platform that technology professionals use to understand how customers interact with their web and mobile products.

As vice president of platform engineering with Fullstory, Kuebrich leads a team of developers who build and manage applications via Google Cloud Platform. When they first created Fullstory’s platform, they used Google’s App Engine to take a no-code approach without requiring on-premises infrastructure. Today, they use Google Kubernetes Engine to automate the deployment of containerized applications.

Click the banner below to read the full CDW cloud computing report.

They also incorporate other Google tools such as Bigtable and BigQuery for data storage and data warehousing, respectively.

Fullstory’s platform engineering capabilities would be much more limited if it didn’t have access to a suite of services designed to help the developers work efficiently, he notes. Add easy scalability and stringent security to the built-in benefits of the cloud, he says, and the choice is clear.

“It’s not that we don’t need to have talented engineers; we definitely do,” Kuebrich says. But cloud-based solutions allow Fullstory’s developers to focus their efforts on higher-level tasks. “Our core competency isn’t running Kubernetes, it’s providing behavioral data to our customers.”

RELATED: How to harness your data and contain costs in the public cloud.

How To Reduce Friction in the Software Development Process

Considered the leading edge of software development, platform engineering combines traditional DevOps processes with advanced techniques and tools to ensure code is written efficiently and securely. The approach is gaining ground among developers due the growing complexity of the trade. Within two years, 80% of large software engineering firms will have dedicated platform engineering teams, Gartner notes.

Also driving the trend is the growing popularity of cloud development environments. Platformengineering.org, an online community for DevOps experts offering training, certifications and professional resources, explains in a 2024 state of the industry report that CDEs use automation and standardization to give platform engineers “a central place to manage the setup, troubleshooting, updating and security” of the software development environment.

“Without a cloud infrastructure, you almost can’t do platform engineering,” says Jim Mercer, program vice president for software development, DevOps and DevSecOps at IDC. “It’s not impossible, but it’s very difficult to do effectively because you need cloud tools for automation.”

These tools include features that help development teams “deal with the extra overhead and friction that come with building intelligent applications,” Mercer says. Some cloud solutions use generative artificial intelligence to produce code.

Reducing friction in the development process is something Kuebrich says he thinks about a lot in his work at Fullstory. It’s also another top reason his team is leveraging Google Cloud.

“The services we get on demand, all we do is click, and it works,” he says.

Working in the cloud also means his team can instantly adjust its platform’s storage and processing capabilities to meet changing demand from the company’s customers. On big shopping days such as Black Friday and Cyber Monday, for example, surging traffic on retail company websites means their systems have to work overtime.

“If we were forced to provision for peak capacity all the time, we wouldn’t be very happy,” Kuebrich says. “But with the cloud, it’s just metered usage consumption. We pay for what we need.”

EXPLORE: Six key components you need for a successful cloud migration.

How Cloud Helps Developers Build an Analytics Pipeline

Another software developer closely minding company spending is David Perez, principal gameplay engineer at Notorious Studios. Launched in 2021 by a team that previously produced games such as Overwatch and World of Warcraft, Notorious is working on its first product, an immersive fantasy action game called Legacy: Steel & Sorcery.

Perez says that he and his colleagues would likely benefit from having some of their systems on-premises. Bigger studios typically take a hybrid approach to platform engineering, he explains: They leverage a mix of tools, some accessed through cloud services, and some developed internally for their specific games.

“We’re a pretty small group, so our entire infrastructure is in the cloud,” Perez says. “You choose what you can with the budget that you have.”

For Notorious, that’s meant working in Amazon Web Services, which it uses primarily for play testing. Leveraging AWS tools, developers built an analytics pipeline for tracking and processing player data.

Gameplay data — such as the number of adversaries killed per game, or where a player died and who killed them, for example — is automatically fed into an AWS data lake and analyzed for valuable insights.

The Notorious designers create dashboards they use to visualize the findings, then they go back to refining the game.

“It’s about seeing what works and doesn’t work and figuring out what to tweak to make it better,” Perez says.  

Perez adds that his team built the pipeline in only two days and deployed it in less than 10 minutes. Amazon services, including the AWS Cloud Development Kit, AWS Glue and the unified business intelligence solution Amazon QuickSight, were critical to the process.

Echoing Kuebrich at Fullstory, Perez says that cloud-based platform engineering tools make his development work easier. But looking ahead, he’s most excited about the flexibility the cloud offers.

When Legacy: Steel & Sorcery is released, Perez explains, Notorious will need server and storage capabilities well beyond what it has now. “It’s nice to have a scalable solution,” he says. “Over the long term, that could really make a difference.”