Data Scientist | Remote

Role Overview

Contract Data Scientist (Remote)

You’ll design and author real-world benchmark tasks grounded in professional workspace materials to evaluate how well language models understand technical documentation and follow precise instructions.

Responsibilities

• Author complex, multi-step benchmark tasks using real workspace files such as:
  • technical specifications
  • architecture documents
  • API references
  • codebases
• Design tasks that test:
  • reasoning over technical documentation
  • instruction following
• Define ground truth outputs for each task and create an objective evaluation rubric
• Incorporate web search and code execution into task design to mirror real workflows
• Contribute to building benchmark datasets for LLM evaluation in the Technology domain

Requirements

• Strong hands-on experience in software engineering, data science, or analytics
• Experience authoring and/or evaluating technical documentation (architecture docs, API references, codebases)
• Experience designing structured multi-step tasks with measurable outputs and evaluation criteria
• Proven ability to work independently in a remote, self-directed contract environment

Contract Details

• Minimum commitment: 15–20 hours/week
• Pay: $90–$110/hour (weekly payments via Stripe or Wise)

How to Apply

• Submit your application via Mercor’s platform
• Complete the screening/assessment as directed
• Engage as an independent contractor upon selection