AI Infrastructure as Part of CI/CD for Mobile Game Development
summary
Mobile-game CI/CD at Player handles C++ engines with ~2 million lines and ~500 developers, ~100M MAU, and six‑week releases, so long builds and store reviews block velocity. A custom AI pipeline uses GitHub Actions runners, Cloud Code, Amplitude and Sentry to run in‑agent incremental builds, apply suggested patches with a cap of three iterations, and use binary search to find failures among 100 pull requests per day; early results cut debug cycles and improve adoption but not yet full trust or automatic commits.
AI-written summary · about
chapters (10)
- 0:00AI in mobile game CI/CD
- 3:43Mobile game CI/CD challenges
- 11:21Measuring CI/CD success
- 13:32Humans slow CI/CD
- 17:11Developer productivity with AI
- 26:08AI-assisted code review
- 37:14AI fixes PR build failures
- 40:51Nightly builds with AI
- 45:51Results and future improvements
- 56:04Q&A
related talks
20mCI/CD Infrastructure for Efficient Mobile Game Development
21mAI-driven optimization of CI/CD pipelines
44mTaking CI and automated testing seriously
35mBeyond the Hype and Hyperbole
18mRe-imagining mobile gaming studios with Gen AI
38mOne Android Dev, Ten AI Teammates: Scaling Delivery With Claude Code
29mHow Agentic AI Is Reshaping Mobile App Quality
30mScalable and Cloud-Native Mobile Game CI/CD Environment Using Unity