Software Engineer
- Context
- Online game company. A distributed Scrum team working on game titles and internal developer tools
- Worked with
- Product owner, team lead, QA, engineers, the framework team
- Period
- May 2024 – Jun 2025
- Stages
At Play'n GO, my main work was refactoring legacy game code for updates to an in-house UI framework, fixing bugs, and developing features for an internal deployment tool. I also helped improve how the team worked: defining what engineers should check before handing work to QA, documenting game-specific integration knowledge, and making actions and errors clearer in the parts of the internal tool I worked on.
The pre-QA checklist
Situation
Engineers did not share a clear understanding of what they should check before handing changes to QA. Some fixes reached QA with problems that a basic run could have caught. Others had already been checked, but engineers and QA disagreed about where their responsibilities began and ended. A "Dev testing" status was added to tickets, but its scope was never defined, and the same disagreements kept returning.
What I did
I proposed that the team define the checks together. I asked the team lead for an initial draft and used it in a working session to discuss what belonged to development and what should remain with QA. I documented decisions as we made them, and we finalised the checklist in a second session.
The checks covered launching a game, loading animations, settings, basic gameplay and sound, as well as common framework-specific failure patterns. Longer test runs remained with QA. My aim was not simply to create a document, but to establish a shared understanding of what was expected.
Result
The checklist became part of the normal development process. Engineers could show what they had checked, QA knew what to expect, and the team lead could explain the agreed scope to the product owner. The recurring friction stopped.
Framework integration and documentation
Situation
Our team integrated updates to an in-house UI framework across a portfolio of around 300 game titles. The titles differed in structure, behaviour and age, so major updates could cause games to freeze or behave unexpectedly. Solutions often depended on the individual game, and much of that knowledge remained with experienced engineers or in chat conversations rather than being consistently documented.
What I did
I refactored legacy game code using TypeScript, Java and object-oriented design, fixing bugs and improving maintainability as I integrated framework updates. I also documented integration steps, known issues, workarounds and code examples, and shared them in a place accessible to my team and other teams.
Internal developer tool
Situation
An internal tool for bulk deployment was useful but difficult to navigate. Its many coloured buttons gave little indication of which actions belonged together or what would happen when they were pressed. Engineers often explained the steps to each other verbally. Tickets could start as broad ideas without detailed designs or specifications, and messages in the built-in terminal were not always clearly identifiable as errors.
What I did
I developed UI features using Vue 3, Electron, TypeScript and Vite, including the flow for selecting and configuring deployment targets. I broke broad tickets into smaller tasks and worked on bug fixes and maintenance. Within the parts I was responsible for, I made terminal errors easier to recognise and added short explanations of what buttons would do.
I also facilitated a team retrospective, drawing on my CSM and CSPO knowledge.