Data Scientist

This role analyzes in-game data from consumer game titles, including examples such as Sonic and Puyo Puyo, to support product and service improvements. Responsibilities include building and validating models, analyzing player behavior, evaluating the effects of initiatives, and turning analytical results into recommendations.

The position uses Python for data analysis and for designing and improving machine-learning models. It also involves statistical modeling, prediction models, SQL-based data extraction and transformation, and improvements to the analytics foundation in collaboration with engineers. Practical experience with JOIN, GROUP BY, subqueries, and other data-shaping methods is required.

Applicants must have at least two years of practical experience in Python-based data analysis or machine-learning model development, and at least two years of practical SQL experience in data extraction and transformation. Experience with one or more of pandas, scikit-learn, LightGBM, or PyTorch is requested, along with foundational knowledge of statistics and data analysis, including regression, classification, and statistical testing.

The role values the ability to identify the essential nature of a problem, reason backward from objectives, communicate opinions clearly, and move coordination forward. It also calls for an interest in learning new analytical methods and technologies, a player perspective, and the ability to interpret what makes games engaging through data. Applicants are expected to enjoy consumer games and play at least once a week.

Preferred experience includes designing model validation, A/B tests, or optimization initiatives; developing BI tools such as Qlik Sense or Redash; and an interest in data engineering areas such as ETL and data pipelines.

The work location is in Shinagawa, Tokyo. The listed employment status is regular employee, although a contract-employee offer may be made depending on skills and experience. The annual salary range is JPY 5,000,000 to JPY 13,000,000.