How EPI Company SE optimized API performance and security testing with Gatling

Why they needed Gatling: Securing the user experience with load testing

To meet their ambitious goals, EPI adopted Gatling Enterprise Edition early on and began testing complex flows:

• API authentication and authorization (JWT, mTLS, HMAC)
  
• Mobile performance under real-world latency
  
• Client certificate validation and signature verification
  
• High-volume transactional traffic was simulated using private Gatling Enterprise runners hosted in a separate AWS region (Paris).

EPI’s teams are structured around features, with each team owning its own test scenarios. A central regression testing suite ensured consistency, while performance metrics like p95/p99 latency and failed transactions were closely monitored against internal SLOs.

The challenge: When APIs meet CPU usage

As their test library grew, EPI encountered a common performance testing challenge: simulating secure APIs came with a high compute cost. In particular, on-the-fly key generation in test scripts led to spikes in CPU usage, skewing results and inflating infrastructure costs.

They also faced friction in their early usage of Gatling Enterprise Edition:

• Test runners consumed more credits than expected
  
• Regression test results lacked centralized storage
  
• The Community Edition made it difficult to maintain a historical view of performance trends
  

It became clear they needed more control, better visibility, and tighter integration with their internal stack.

The solution: Scaling smart with Gatling Enterprise Edition

EPI’s engineering team migrated fully to Gatling Enterprise Edition to address these growing pains. This gave them the tools to scale without compromising efficiency or governance:

• Kotlin-based scripting, matching their backend language, made it easier for developers to write and maintain performance tests
  
• Private Locations on AWS reduced data egress, improved test stability, and gave them more control over cost
  
• Optimized script design lowered CPU usage and reduced the need for multiple test runners
  
• Enterprise Cloud reporting enabled centralized storage, versioning, and integration with Confluence
  
• SSO and Datadog integrations brought performance testing in line with EPI’s broader security and observability frameworks
  

Gatling also offered small but meaningful UX upgrades—like simulation tab customization—that made test results easier to consume across teams.

“What I like about Gatling is that it lets us write performance tests in Kotlin, the language we use every day. That really lowers the barrier for developers to contribute.”

Ulrich Winter, Team Lead Infrastructure & Technology

The results: A repeatable testing culture

With Gatling, EPI didn’t just improve their tooling—they operationalized performance testing as a routine part of development. Today:

• About 15 engineers across multiple teams collaborate on tests
  
• Manual regressions run every two weeks, with additional tests triggered by feature changes
  
• Security flows are validated regularly, including client certs and signature handling
  
• Plans are underway to automate runs via GitHub and expand CI/CD integration
  

“Gatling has helped us execute regular load tests efficiently and consistently compare results across runs.”

Ulrich Winter, Team Lead Infrastructure & Technology

Thanks to Gatling Enterprise Edition, EPI Company:

• Removed bottlenecks from cryptographic test cases
  
• Reduced test infrastructure costs via optimized execution
  
• Gained shared ownership across engineering teams
  
• Aligned testing workflows with DevSecOps goals
  
• Delivered a high-trust platform, backed by continuous performance validation

‍