Seeking Insights on Test Orchestration & Selection in CI

We’ve got a growing UI automation suite (Playwright + some Selenium) wired into CI. Right now we are doing mostly basic orchestration using GitHub Actions and some scripts to split tests. Also, most PRs still run a big chunk of the suite, so pipelines are getting slow and flaky.

For teams who solved this:

  • Did you build custom in-house orchestration / selection or use a third-party tool?

  • What kind of benefits did you get? Did it actually justify the cost/effort?

  • Side question: did security/compliance ever push back on plugging a third-party service into CI (access to repos, logs, test data, etc.), or was that a non-issue ?