Remote Data Scientist & Quant Analyst for AI Evaluation

24-Mag Llc; New York (NY); Remote; USD 75,768 - 117,096; Full time

Job summary

24-MAG LLC is seeking a highly skilled data scientist or quantitative analyst for a full-time remote consulting role. You will design advanced data-analysis tasks, develop reproducible notebooks, and perform rigorous statistical comparisons for frontier AI benchmarks.

The role emphasizes transparent reporting, methodological rigour, and collaboration with researchers to refine tasks and scoring rubrics. Approx. 35 hours per week.

Qualifications

• Masters or PhD in a quantitative field is highly relevant and welcome.
• Hands-on data cleaning, correlation, hypothesis testing, and interpretation are essential.
• Proficiency in Python and common data libraries is required.
• Experience with reproducible analysis workflows and notebooks is expected.
• Experience communicating complex quantitative findings to stakeholders is valued.

Responsibilities

• Design realistic, professional data-analysis tasks and multi-step problems.
• Develop reproducible reference notebooks using Python, pandas, and NumPy.
• Conduct thorough statistical method comparisons and evaluate model outputs.
• Review AI-model analyses for statistical accuracy and reporting quality.
• Collaborate with researchers to refine tasks, notebooks, and rubrics.

Skills

Data cleaning

Statistical correlation

Hypothesis testing

Python

pandas

NumPy

Jupyter Notebook

Google Colab

Git

Reproducible workflows

Communication of results

Attention to detail

Available 35h/week

Education

Master's degree in statistics/data science or related

PhD in quantitative field

Tools

Jupyter Notebook

Google Colab

Git

Job description

24-MAG LLC is seeking a highly skilled data scientist or quantitative analyst for a full-time remote consulting role. You will design advanced data-analysis tasks, develop reproducible notebooks, and perform rigorous statistical comparisons for frontier AI benchmarks.

The role emphasizes transparent reporting, methodological rigour, and collaboration with researchers to refine tasks and scoring rubrics. Approx. 35 hours per week.