My Contributions

Conduct regular smoke tests. Create Jira tickets for new bugs, record recurring bugs, and update existing tickets that presented altered behavior. My work involved completing run-throughs, recording feedback - further elaboration in section marked "End-user Feedback" -, updating test suites, and conduct targeted analysis based on bug reports. In addition to Quality Assurance work I also conducted a final internship project further elaborated on in the section and subsections marked "Final Project".

End-user Feedback

With Closer the Distance and Goodnight Universe we participated with late-stage development preparing for release. Therefore, we were often responsible for feedback on final polish in addition to our standard QA services. We approached this by maintaining a holistic view of the projects and recording memorable moments, characters, and genre comparisons. I would later report my experience back to my manager who acted as point of contact with the studio and Skybound leadership. These sessions were usually formatted as open conversation focused on specific points in the story, touching on each independently and in-depth.

Final Project

For my final project I researched insights regarding feature considerations for Invincible VS from the Fighting Game Community (FGC) and related titles. After my initial research I compiled my findings and presented to studio leadership, fielding any questions.

AI Sentiment Tracking

Using SteamDB I was able to collect extensive user reviews, wish-list data, sales projections, and player data. Although there are several AIs proficient in sentiment tracking I landed on GPT-4o because it is one of the most high scale, well trained LLMs. If I were to revisit this project today I would likely revisit this choice as many LLMs have begun to specialize into sentiment tracking.

Using GPT I was able to parse through reviews to find only those mentioning features that were included in my study and then to distill the sentiment towards that feature. This provided the study with a much more precise estimation of the impact of specific features on games reception rather than relying on the overall rating associated with each review which dramatically improved the granularity and construction time of my report.

Looking Back

• Providing Qualitative Feedback: Prior to this role I primarily had experience in more technical QA so providing polish feedback was a new experience. To address this I ensured that my note taking methods included both positive and negative sections which I would address at the end of each act. I would record feedback as a brief stream of thoughts in the moment and would revisit my notes at the conclusion of each project to condense, extract common themes, and draw parallels to create feedback that was easier to digest.
• Stand out Stand up: At the conclusion of my project I was given the opportunity to pitch how I would imagine a feature. Although it was a spur of the moment conversation, many of the senior designers were aligned around two different methods. Instead of presenting a third method, I addressed the current ideas and which parts I believed to hold great value according to my research. In hindsight, it was my opportunity to present something new, discussed how it aligned with my research, and leave it to the team to decide which parts were of value to the project.