[Contract] PhD-Level Computational Biologist - Design Data-Centric Benchmarks for AI Models

[Contract] PhD-Level Computational Biologist - Design Data-Centric Benchmarks for AI Models

Mercor

Remote

Role Overview

Mercor is seeking computational biology experts to contribute to a unique project with a top-tier AI research organization. This short-term initiative challenges AI models with hidden-answer “mystery problems” grounded in real biological data. Experts will design, validate, and anonymize complex datasets to evaluate model reasoning capabilities, not lookup accuracy. This is a compelling opportunity for data-driven life scientists to shape next-generation AI evaluation benchmarks.

Key Responsibilities

• Design biologically grounded problems with a single correct answer and ≥20 plausible distractors
• Identify, download, and preprocess datasets from public repositories (e.g., GEO, SRA, NCBI)
• Subsample and anonymize datasets to prevent metadata-based inference
• Independently validate solutions through custom analysis pipelines and visualizations
• Operate within technical constraints (e.g., file size, task time limits)

Ideal Qualifications

• Hands-on experience analyzing sequencing or omics datasets (e.g., RNA-seq, WGS, mass spec)
• Proficient with bioinformatics tools and formats (e.g., BLAST, samtools, DESeq2, FASTA/FASTQ/BAM)
• Comfortable coding in Python, R, or bash, and working in Jupyter notebooks
• Understanding of biological experiment design and lab-to-data nuances (e.g., batch effects)
• Advanced degree (MS, PhD) or research background in computational biology, genomics, or bioinformatics

More About the Opportunity

• Remote and asynchronous - set your own schedule
• Expected commitment: 15–20 hours/week
• Project duration: ~1 month
• Potential for additional projects based on performance and interest

Compensation & Contract Terms

• $65–85/hour depending on experience and geography
• Paid weekly via Stripe Connect
• Structured as a freelance contract - independent contractor status