AI Engineer Intern (Model & Product Evaluation)
The AI Product team focuses on building practical AI solutions that improve how people work and interact with our products. Our work spans both internal and external projects, including AI-powered operational tools, AI agents, and integrations across regions and games, supporting users around the world. The team values hands-on execution and fast experimentation (test, learn, move on) to build AI products that serve users more meaningfully and effectively.
In this role, you will identify areas where our AI products can be improved, understand the underlying issues, and work closely with product and ML teams to prioritize and implement solutions. You will play an important role in making our AI products more reliable, useful, and effective in meeting users’ needs. So if you are someone who enjoys solving problems, exploring new ideas, and taking a hands-on approach to experimentation, this role and team will be for you.
- Support in building and improving an evaluation algorithm that learns from past human decisions to assess the quality of new AI suggestions and identify when human review is required.
- Turn evaluation findings into testable hypotheses to improve an AI platform, such as its prompts, knowledge sources, decision logic, or model configuration, and run experiments to validate their impact.
- Drive failure case analyses, and evaluation reports to track improvements in quality over time.
- Work closely with CS leaders, ML engineers, product managers, and other stakeholders to translate reviewer feedback into prioritized experiments and system improvements.
- Bachelor’s or Master’s student (final-year preferred) or recent graduate in Computer Science, AI, Mathematics, Data Science, Statistics, or a related field.
- Understanding of statistics, including agreement measures and sampling.
- Strong sensitivity to AI product experience, with the ability to clearly describe observed issues and their possible causes.
- Clear logical thinking, with the ability to organize large volumes of samples into reusable issue frameworks.
- Open-minded, agile, and proactive, with a strong willingness to learn.
- Basic ability to build quick working prototypes using AI coding tools such as Claude Code or Cursor; a portfolio or side project is a plus.