Collaborative Quality at the Core: AI-Powered Unit Test Synergy to Reduce E2E Dependency

Unit tests are critical pieces of testing framework such as the Testing Pyramid, and they're widely used in the software industry. However, many organizations treat unit testing as a separate, reserved developer's domain, where a lack of synergy with QA is the default habit, resulting in over-reliance on end-to-end (E2E) tests, costly test maintenance, and inefficient feedback loops.

I want to share how my team integrated unit tests into a shared testing strategy involving developers, Product Owners and QAs.

By improving collaboration and aligning test coverage early in the development process, we can identify which scenarios are covered by unit tests and which need more focus from E2E automation, reducing unnecessary E2E tests, their scope, and the efforts for their maintenance. Additionally, we leveraged AI-powered tools to identify and review unit tests more efficiently, further streamlining the testing process and helping focus on areas not covered by unit tests, improving overall test efficiency.

Our approach contributed to measurable improvements in key DORA metrics, such as a higher Deployment Frequency (DF) and a reduction in Change Failure Rate (CFR).

In the talk, I'll present data on the results, challenges, trade-offs, and lessons learned

Paper | Presentation

Simone Colosimo

Simone was born and raised in Rome, Italy, and graduated in Telecommunication Engineering. He moved to France in 2007, and since then, has worked in the quality domain, first in the telecom world, then bifurcated towards the software development field to broaden his horizons.