Apply data science and Python expertise to evaluate coding agents through reproducible environments, testing, and rigorous solution verification. This Competitive Data Scientist role focuses on building reliable tasks and reference solutions for evaluating coding agents. You will prepare relevant data and dependencies, investigate failures, and refine tasks from observed behavior. The work suits candidates with data science, machine learning, quantitative modeling, or research competition experience. Key Responsibilities:

Build reproducible environments, data, dependencies, and reference solutions Develop tests distinguishing correct and flawed approaches Run coding agents against tasks and investigate failures Identify implementation errors, assumptions,



and data leakage Document requirements, reasoning, and verification steps Requirements:

Experience in data science, machine learning, or quantitative modeling Relevant research competition experience, including Kaggle or DrivenData Solid problem-solving ability and methodological understanding Programming ability to implement and verify solutions Python testing, debugging, and unfamiliar-code experience preferred Bring analytical judgment, practical programming, and evaluation expertise to strengthen coding-agent task quality and correctness.

#J-18808-Ljbffr

📌 Competitive Data Scientist Evaluator (Winnipeg)
🏢 Great Value Hiring
📍 Winnipeg