- Experience
- 3+ yrs
- Location
- United States, Canada, United Kingdom, Ireland, Australia May Also Be Accepted
- Work mode
- Remote
- Job type
- Contract
- Duration
- 12 Weeks
- Weekly
- 10+ hours
- Salary
- $40–$120 / hour
- Posted
- today
About the role
The Life Sciences AI Evaluator – Computational Biology & Bioinformatics will evaluate frontier AI models on realistic biological data analysis tasks, including genomics, transcriptomics, variant interpretation, multiomics, statistical genetics, and machine learning applied to biology. The role involves designing expert-level scenarios, running them through AI agents, comparing outputs, creating grading rubrics, and documenting analytical errors against professional standards. Candidates must have at least 3 years of hands-on biological data experience, an in-progress Bachelor's degree or higher in a relevant field, strong scientific judgment, and excellent written communication.
Key Responsibilities
- Design challenging, realistic computational biology tasks based on day-to-day professional workflows, including the scenario, prompt, supporting files, and expected analytical standard.
- Create tasks involving differential expression, RNA-seq count matrices, variant prioritization, batch effect diagnosis, pipeline audits, single-cell workflows, multiomics integration, and related biological analyses.
- Assemble or adapt realistic supporting materials such as count matrices, sample sheets, metadata, VCFs, pipeline logs, quality control outputs, reference annotations, notebooks, correspondence, and reports.
- Run designed tasks through frontier AI agents and evaluate the resulting analysis reports, annotated tables, figures, notebooks, code, and interpretations against professional standards.
- Compare model outputs generated from identical prompts and files, determine which performed better, and document the strengths, weaknesses, and failure modes of each response.
- Write detailed grading rubrics defining the requirements for a correct analysis, including appropriate normalization, multiple testing correction, statistical reasoning, reproducibility, and biological interpretation.
- Flag concrete analytical failures with supporting evidence, including ignored batch effects or confounders, inappropriate statistical tests, uncorrected p-values, incorrect references or annotation versions, silently dropped samples, fabricated results, and code that does not match its narrative claims.
- Evaluate whether model responses remain faithful to the requested task and identify interpretations that are off brief, scientifically unsupported, or inconsistent with the underlying data.
- Contribute expertise across genomics and variant interpretation, bulk and single-cell transcriptomics, proteomics, metabolomics and multiomics, population and statistical genetics, clinical genomics, metagenomics, phylogenetics, and machine learning applied to biology.
- Review and refine tasks created by other experts, incorporate accurate feedback, and independently investigate unfamiliar questions or incomplete instructions when needed.
Required Qualifications
- At least 3 years of hands-on experience analyzing real biological data in industry, an academic lab, or a research institute; academic lab and research institute experience counts after undergraduate education.
- An in-progress Bachelor's degree or higher in Biology or a directly related field, including molecular or cell biology, genetics, immunology, neuroscience, biochemistry, bioinformatics, or computational biology, completed in the U.S., Canada, Europe, or the UK.
- Directly related education in biostatistics, computer science, or statistics is also accepted when paired with substantial biological data experience.
- Ability to write, debug, and explain your own analysis code in Python and/or R.
- Depth of experience in at least one relevant area, such as genomics and variant interpretation, bulk or single-cell transcriptomics, proteomics, metabolomics or multiomics integration, population and statistical genetics, machine learning applied to biology, clinical genomics, metagenomics, or phylogenetics.
- Working understanding of several adjacent biological data-analysis areas and their workflows.
- Independent ownership of a multistep analysis from raw or messy data through cleaning, processing, and final interpretation.
- Familiarity with reproducible research practices, including notebooks, scripts, version control, and workflow tools.
- Ability to assess normalization choices, batch effects, confounding, multiple testing correction, statistical power, and biological interpretation.
- Ability to draw on real-world experience and day-to-day workflows to create scenarios that test whether an AI system can perform biological data analysis correctly.
- Current or recent hands-on bench or analysis practice at an individual-contributor level, rather than solely managerial experience.
- Full professional or native-level written and spoken English, strong written communication, and the ability to explain complex scientific reasoning clearly and concisely.
- Ability to explain why a result is wrong, not only identify that it is wrong.
- Comfort with ambiguity, strong attention to detail, and the ability to verify document claims against underlying data.
- Ability to interpret feedback, determine which feedback is correct, apply it independently, and investigate answers when blocked.
- Ability to ramp quickly on unfamiliar work from written material and incomplete instructions.
- General familiarity with AI and large language model tools, including professional use of Claude or ChatGPT in life sciences work.
- Baseline technology literacy, including cloud file tools, browser profiles, desktop applications, file conversion, and file sharing.
- Availability for at least 10 hours per week, with no weekly maximum.
- Based in the United States, Canada, or the UK; Ireland and Australia may also be accepted.
Benefits and Perks
- Fully remote work environment.
- Flexible, self-directed long-form work.
- Opportunity to contribute to the development of better artificial intelligence systems for life sciences.
- Minimum availability of 10 hours per week with no stated weekly maximum.
- Consistent availability is valued, and additional hours are welcome.
Skills
- Differential expression
- Variant interpretation
- Single-cell workflows
- Multiomics integration
- Statistical genetics
- Computational biology
- Bioinformatics
- RNA-seq
- Quality control
- Variant prioritization
- Batch effect diagnosis
- Pipeline auditing
- Genomics
- Bulk transcriptomics
- Single-cell transcriptomics
- Proteomics
- Metabolomics
- Population genetics
- Machine learning applied to biology
- Clinical genomics
- Metagenomics
- Phylogenetics
- Analysis code
Education
Required
- Bachelor
Selection process
Step 1
Submitted
Someone who fits the role and has told you they are interested, with their resume.
Step 2
Client interviews
Step 3
Selected
If they qualify and are selected, your payout is due.