Functional QA, live operations and early ownership across events and testing teams.
Live Ops TesterAugust 2019 – December 2020 · India
500+
defects personally reported
~7
live events led in QA
~5
testers mentored
4
titles affected by a clipboard issue I identified
The day-to-day
Test what happens when people actually play.
I tested released games end to end from the player’s perspective: gameplay, functionality, internet connectivity, live events and reward distribution.
The role also included community issue investigation, test plans, regression, mentoring and reporting. Alongside live titles, I supported selected Alpha and Box builds for unreleased games.
My focus was functional behaviour and player experience. The practical question was whether the game and its connected systems behaved as intended under the conditions being tested.
Defect investigation
A report can begin with very little information.
When I found an issue
I retraced the actions, narrowed the reproduction steps, captured the build number and supporting screenshots or video, and wrote a PAL summary: Problem, Action, Location.
I assigned probability, reproduction and severity ratings, then routed the completed report to the appropriate functional point of contact.
When a player found an issue
Reports on Reddit, Discord and other public forums sometimes offered only an image or a short video. I examined the evidence, formed hypotheses about the affected systems and attempted reproduction.
Once verified, I turned the player’s report into a complete defect developers could investigate.
Retest beyond the original path.
For high-severity, high-priority or high-probability fixes, I confirmed the original reproduction path and spent an additional five to ten minutes trying alternative paths. If another cause surfaced, I documented it and involved the relevant point of contact before closure.
Event ownership
Trusted early. Responsible for the details.
2019Hungry Dragon · Thanksgiving event
About three months after joining, I was entrusted to lead QA for Hungry Dragon’s Thanksgiving event. It was an early expansion of responsibility.
Across approximately seven events, I typically coordinated around three additional testers. I reviewed event documentation in advance, defined what was in and out of scope, answered questions and acted as the QA point of contact.
I created test plans and cases, tracked progress and shared end-of-day reports daily during events I led. During normal operations, reporting was twice weekly.
Updates went to QA and department leads, producers and wider management so testing progress and concerns were visible.
Quality with consequences
Look beyond whether a test case passes.
01
Clipboard behaviour
I identified unintended clipboard access affecting four mobile titles, recognised a potential privacy and platform-compliance concern and escalated it through QA. Development teams fixed the issue.
02
Readiness recommendations
On two occasions, while supporting The Crew Motorfest and Rainbow Six Mobile, I formally recommended that features needed rework. Final go/no-go authority remained with leadership.
03
Player confusion
When several testers independently perceived intended behaviour as a bug, I raised the potential usability problem in reporting. The same confusion could affect players.
Review the work. Support the person.
I mentored approximately five testers and reviewed their defect reports. Urgent corrections were made where needed, with feedback shared privately so reporting quality could improve without holding up critical communication.
Day-to-day testing priorities were set through Agile-style meetings with the QA lead. I aligned work to current events, releases and live-operations needs.