Bioinformatics AI Evaluation Task Designer

You will create challenging computational genomics and bioinformatics problems for advanced AI evaluation. Each task should require an AI system to interpret biological data, write and run code, use scientific file formats, and produce outputs that can be checked objectively.

The role combines scientific judgment, computational analysis, benchmark design, reference-solution development, and detailed quality review.

• Design novel tasks using FASTA, FASTQ, VCF, BAM, SAM, BED, TSV, and CSV files.
• Work with sequence annotations, expression data, and germline or somatic variant data.
• Create task specifications, input datasets, expected output schemas, deterministic ground truths, and expert reference solutions.
• Build reproducible workflows with Python, Linux command-line tools, and established bioinformatics libraries.
• Check scientific correctness, solvability, reproducibility, and difficulty.
• Create grading criteria that separate scientifically correct answers from plausible-looking but incorrect results.
• Document deliverables and communicate technical requirements, blockers, and progress to project leads and reviewers.

This is a remote contractor assignment lasting five weeks. The work requires strong scientific and programming experience in bioinformatics or a closely related field.

• Pay: $150 per approved task.
• Time: At least 20 hours per week.
• Schedule: Include a four-hour overlap with Pacific Time.
• Location: Remote and open worldwide.
• Education or experience: Ph.D., postdoctoral experience, or equivalent research experience in bioinformatics, computational biology, genomics, computational genetics, or a related discipline.
• Required skills: Strong hands-on Python programming, experience with biological sequence or genomics datasets, and comfort with Linux and command-line environments.
• Scientific knowledge: Understand the experimental and biological context behind computational analyses.
• Helpful experience: Variant analysis, transcriptomics, sequence analysis, phylogenetics, population genetics, functional genomics, or clinical genomics.
• Helpful tools: Biopython, pandas, NumPy, SciPy, samtools, bcftools, BLAST, PLINK, or equivalent tools.
• Also useful: AI or language model evaluation, benchmark or automated grader design, reproducible pipelines, Docker, or containerized scientific workflows.
• Benefits: The assignment does not include medical or paid leave.

OpenTrain is where you start the application for this AI training role. Apply on OpenTrain with your resume, then complete the application on the hiring site.

AI training work uses examples, evaluations, and human feedback to help artificial intelligence systems produce better results. In specialist projects like this one, researchers and technical experts help ensure that scientific tasks, reference answers, and grading rules are accurate.