- Phase 1: Empowering developers with seamless PR workflows
- Phase 2: Enabling infrastructure observability and SOC 2 Compliance
- Phase 3: Expanding visibility and set the stage for developer productivity insights
Introduction
Since it was founded in 1999, SurveyMonkey has evolved into a global provider of online survey and experience management solutions, serving millions of users worldwide including 95% of the Fortune 500.
But being a mainstay in the space comes with its own set of unique challenges ā namely a significant amount of technical debt that SurveyMonkeyās engineering team has worked to modernize.
When Bart Dziekan first joined SurveyMonkey, he saw an opportunity to rebuild the companyās infrastructure from the ground up. As Head of the Cloud Solutions team, Bartās primary goal was to create a modern, scalable developer platform that would not only simplify deployment but also support faster development cycles and cleaner workflows.
āThat's one of the reasons I came over to SurveyMonkey. You don't get the opportunity to start fresh at many places, but here I was given the chance to take all my background knowledge and my vision of what an ideal infrastructure looks like and start from scratch. LinearB workflow automation was instrumental in standardizing all our new processes.ā
Bart Dziekan
Head of Cloud Solutions, SurveyMonkey
Now, three years later, SurveyMonkeyās infrastructure is nearly unrecognizable. Manual EC2 instance rebuilds have been fully phased out. In their place, the SurveyMonkey engineering team now operates within a Kubernetes-based environment, leveraging a GitOps-driven platform that employs LinearB workflow automation and standardized pull request workflows.
SurveyMonkeyās journey with LinearB followed a three-phase rollout:
Phase 1: Empowering developers with seamless PR workflows
When Bart started at SurveyMonkey, infrastructure practices varied widely. Developers had near-complete freedomābut that came at the price of inconsistency. Bart headed up a company-wide migration from AWS to EKS, Kubernetes, and Terraform, which inherently required tighter controls. As a result, PR standardization became essential to safeguard deployments and ensure consistency across teams.
To better enforce approval rules across deployment environments, SurveyMonkey leveraged open-source tools like Atlantis for Terraform ā but Atlantis required manual PR approvals, which created a bottleneck for developers in Dev and Test environments.
Thatās when Bart discovered LinearB workflow automation. With its YAML-based policy-as-code capabilities, LinearB allowed his team to implement highly customizable PR automation rules. They auto-approved safe changes based on folder structure (e.g., Terraform or Argo-related code), bypassing the manual approval bottleneck in Dev and Test, while maintaining human review in production for compliance.
"We applied LinearB automation to define folder-specific rules, label PRs, and even route break-fix changes through specialized workflows. Now, if a specific label is being used for a PR as a break-fix, that creates a Slack message in one of our incident rooms."
As a result, SurveyMonkey developers move faster, theyāve noticed improvements in Cycle Time, and the infrastructure team has stopped fielding complaints about approval delays. Instead of circumventing rules, Bartās teams now trust the process.
Phase 2: Enabling infrastructure observability and SOC2 Compliance
But Bartās mission wasnāt just to modernize infrastructureāit was to create a resilient and auditable platform for application teams to build upon.
"The goal was to give developers the ability to build great software without worrying about infrastructure. They should be able to push their code and trust that the system will handle the rest."
By migrating from legacy tools like EC2 and Koala (an internal Ansible wrapper) to Kubernetes and GitOps, Bart and his team offloaded the infrastructure burden from developers. Now, EC2 instances rebuild every 24 hours automatically, removing the need for manual intervention.
"Previously, developers had to rebuild infrastructure themselves if there was a CVE. Now, that responsibility is fully ours. It means more work for my team, but itās worth it because our developers get to focus on what they do best."
LinearB workflow automation also contributed to improvements around SOC2 Compliance and observability. By labeling PRs with contextual information (e.g., "Missing JIRA Ticketā) the team enforced hygiene standards that made future audits and retrospectives easier. When incidents occur, itās now much easier to trace changes back to the source.
āWe keep finding use cases where we're like, āHey, we'll just use LinearB automation for this.ā So it's not been a one-stop shop for us. Every time we see an opportunity to automate low-hanging fruit ā like PR cleanliness for example ā we take it. Sometimes our developers arenāt consistent with putting explanations of what exactly theyāre doing in PR descriptions. But our JIRA is so clean that if it's tied to a JIRA ticket it's super easy to cross reference. LinearB now reminds our devs to link every PR to our PM instance.ā
Some of SurveyMonkeyās favorite workflow automations rules include flagging PRs with sensitive files, unlinked JIRA tickets, and missing tests, so their team never deviates from compliance standards.
As SurveyMonkey scales its CI/CD practices, these layers of standardization ensure that systems are not just fast, but also stable and secure.
Phase 3: Expanding visibility and setting the stage for developer productivity insights
While LinearB solved immediate pain points for SurveyMonkey with PR automation and standardization, it also opened the door for a broader journey with developer productivity insights.
With executive stakeholders increasingly interested in engineering performance and ROI, Bart sees LinearBās Developer Productivity Platform as the next evolution.
"Right now, we donāt have much of a view into which teams have higher velocity, where the bottlenecks are, or why some teams deliver faster than others. Weāre mostly tracking PR open times, which can be a bit reductive."
Bart is exploring how LinearB can provide actionable metrics beyond what JIRA or GitHub alone can offerādata on Cycle Time, Planning Accuracy, and developer WIP. He recognizes that these insights will help SurveyMonkey executives better understand team performance and guide smarter investment decisions moving forward.
"Just like we use observability tools to monitor platform health, we want similar insights for our engineering teams. Weāre excited to see what we can get out of the LinearB platform."
While adoption of these advanced features is just beginning, Bart sees LinearB as a long-term partner in SurveyMonkeyās modernization journey.
The results
In just three years, Bart and his team have radically transformed Survey Monkeyās legacy infrastructure. PR labels and LinearB workflow automation make it easier to trace changes and pass audits. Application teams no longer manage infrastructure. Instead, they focus on building great products. Plus, engineers understand and embrace the need for structure.
"Standards used to be a dirty word,ā Bart says. āNow, theyāre just a part of the way we work."
Moving forward, Bartās vision remains unchanged: empower developers, reduce friction, and build infrastructure that ājust works.ā LinearB is key to realizing that vision.
"We donāt want to build custom tools unless we absolutely have to. LinearB automation gave us everything we needed right out of the box. Every time we run into a new PR problem, we ask: can LinearB solve this too? And so far, itās done exactly what weāve needed."
Bart Dziekan
Head of Cloud Solutions, SurveyMonkey
For Bart and the Cloud Solutions team, this is just the beginning. As they move into Phase Twoāoptimizing CI/CD pipelines, enabling AI-powered insights, and tracking Developer Productivity insightsāthe foundation theyāve built with LinearB has set the stage for their next transformation.
