Competitive Data Scientist for AI Benchmarks

Vettedlance; United States; Remote; USD 83,000 - 165,000; Part time

Job summary

Vettedlance is seeking a data-science and ML challenge designer who creates reproducible task environments, writes reference Python solutions and develops tests to distinguish correct from flawed approaches. You will run agents, analyze failures and tighten leakage or loopholes to improve evaluation integrity.

The role is remote, hourly, independent contractor work with flexible hours (10–40 hours per week). Strong ML practice and clear documentation are essential for success.

Qualifications

• Strong public data-science competition record, or equivalent ML practice.
• Expert Python and ML libraries.
• Rigor about data leakage and evaluation design.
• Clear written documentation.

Responsibilities

• Build reproducible task environments with data and dependencies.
• Implement verified reference solutions in Python.
• Write tests that separate correct from flawed approaches.
• Run agents, study failures and fix leakage or loopholes.

Skills

Python

ML libraries

Data science

Documentation

Job description

Design data-science and ML challenges that coding agents must solve, and write the reference solutions that prove them solvable. You will run agents against your tasks and tighten anything they exploit.

What you will do

• Build reproducible task environments with data and dependencies
• Implement verified reference solutions in Python
• Write tests that separate correct from flawed approaches
• Run agents, study failures and fix leakage or loopholes

What you need

• Strong public data-science competition record, or equivalent ML practice
• Expert Python and ML libraries
• Rigor about data leakage and evaluation design
• Clear written documentation

Contract

• Remote, hourly contract (independent contractor)
• Part-time, flexible: 10–40 hrs/week
• Remote

Pay: $60 – $120 / hour (USD). Paid weekly in USD. Typical rate $60–$120 per hour, confirmed after a short paid assessment.