This session will show how the Playrix QA team improved its processes for Gardenscapes and other games. They started with a small automation prototype and gradually built a unified framework, improved pipelines, and used AI in test design, log analysis, and prototyping. Vitaly will share practical lessons, including what worked, what didn’t, and how to make a QA team more productive.

Join this presentation with Lucas and Tomasz, as they run through how Publishing QA helps keep product quality high at every step of development, reduces publishing risks, and ensures that products made by external partners meet CDPR standards for player experience. Publishing QA plays a critical role in protecting player experience when development is spread across internal teams and external partners, and the session will help attendees take away learnings of how to catch issues early, rather than reacting late. 

This session explores how one team reduced a seven figure external QA spend by 75 percent while continuing to support a live operations game. It looks at how they redefined what QA work actually added value, shifted from outsource heavy models to embedded craft focused QA, and changed day to day testing practices to work more closely with developers. The talk also covers how hiring profiles, test case practices, and production workflows evolved, and where internal and external QA each proved most effective.

- How to identify and remove low value QA activities, from over engineered test cases to duplicated platform coverage

- Why closer collaboration with developers and stronger technical testing skills can dramatically improve efficiency

- How to balance embedded and external QA models to support live projects without inflating cost or overhead

With a range of QA roles represented at the event with DevQA, Engineering, Publishing QA and so on – there are so many forms that go way beyond testing. But beyond different categories of QA, there is also great variety in the QA setup between studios of different sizes. The goal of this panel is to share similarities and differences beyond technical maturity, to understand what a shared learning can truly be, and what capabilities are only possible in a studio of a certain size.

- How is testing different across the studio sizes, what are we capable of?

- Where can we see QA going in the future, is there a path from small to AAA QA setups?

- How can AI be implemented across the different scales of studios?

When we have asked the question of what the future of Localization would look like, almost always, AI is in some way involved in this future. We cannot currently know in what capacity this might be, but what we can do is understand how we can develop the suitable skills to build a career for ourselves that is aligned with a human/AI Localization function. A function that can still champion human authenticity, and game experience.

Join Mike to understand both from a leadership perspective and for someone early in their Localization career, what skills that are not traditional for Localization, would be imperative to stand out and be future-proofed for new era of Localization.

At one point, we grew tired of having ten thousand strings scattered all over the project. So, we made, and open-sourced, a plugin to move text from blueprints and other places to string tables, with good keys and context fields. It also lets us know exactly where each string table entry is used in the game and provides a bunch of QoL improvements. Best of all, you can have it.

AI in localization is often treated as all or nothing, and it can be a difficult topic to know whether to embrace or reject. This session focuses on the in between. Learn about this studios, structured approach to evaluating AI over time, without chasing quick wins or assuming it will become a silver bullet. Drawing on 6 to 12 months of refining this process, the session will present how teams assess what is genuinely useful and share clear guidance internally with localization and production, beyond just conversations about instant translation or AI voiceovers.

Localization teams not only translate, but they have access to massive amounts of language data, like player feedback, voice scripts, cultural context etc. There is huge opportunity in this for Localization teams to not only act as a service to align all of this data so that it is accessible for our studios, but what is to say Localization cannot be the ones to own this information, and analyse this data. Valuable insights to improve game quality and experience can be pulled from this information, but the analysis skill often comes outside of the usual priorities for Localization.

Join this session to learn how these skills can equip teams to diversify the roles and value Loc can bring, in the future of the discipline.