BioMystery Benchmark Specialist

TELUS Digital AI Data Solutions; Philippines; On-site; PHP 7,561,000 - 9,452,000; Full time

Job summary

TELUS Digital is looking for PhD-level Biology and Life Sciences researchers in India to contribute to BioMystery, an advanced benchmark designed to evaluate how well AI systems can reason through complex biological and biomedical problems.

As a BioMystery Benchmark Specialist, you will use your research expertise to create, solve, validate, and evaluate challenging scientific problems that require more than factual recall.

Qualifications

• PhD completed or currently pursuing in Biology, Biotechnology, Biochemistry, Biomedical Sciences, Life Sciences, or a closely related discipline.
• Demonstrated research experience in a relevant biological or biomedical field.
• Excellent written and spoken English.
• Ability to work independently in a remote, project-based environment.

Responsibilities

• Design original, challenging biology and biomedical problems for evaluating advanced AI systems.
• Create problems requiring multi-step scientific reasoning, evidence interpretation, hypothesis evaluation, and domain expertise.
• Develop clear, scientifically rigorous solutions and expected answers.
• Evaluate AI-generated responses for scientific accuracy, logical reasoning, biological plausibility, and evidence interpretation.
• Review benchmark questions for ambiguity, unintended shortcuts, insufficient information, or multiple valid interpretations.
• Provide structured feedback to improve benchmark quality and model evaluation.

Skills

Research experience

Analytical reasoning

English proficiency

Education

PhD in Biology or related

Job description

TELUS Digital is looking for PhD-level Biology and Life Sciences researchers in India to contribute to BioMystery, an advanced benchmark designed to evaluate how well AI systems can reason through complex biological and biomedical problems.

As a BioMystery Benchmark Specialist, you will use your research expertise to create, solve, validate, and evaluate challenging scientific problems that require more than factual recall. The work focuses on testing whether advanced AI systems can connect evidence, interpret scientific information, identify underlying biological mechanisms, resolve competing hypotheses, and reach scientifically defensible conclusions.

This is a research-focused AI evaluation role, not a traditional data annotation position. You will be expected to apply your academic and research expertise to develop high-quality benchmark material and critically evaluate AI-generated scientific reasoning.

What You'll Do

• Design original, challenging biology and biomedical problems for evaluating advanced AI systems.
• Create problems that require multi-step scientific reasoning, evidence interpretation, hypothesis evaluation, and domain expertise.
• Develop clear, scientifically rigorous solutions and expected answers.
• Evaluate AI-generated responses for:
• Scientific accuracy
• Logical reasoning
• Biological plausibility
• Evidence interpretation
• Missing or incorrect assumptions
• Contradictions and hallucinations
• Identify subtle scientific errors that may not be apparent through simple factual checking.
• Assess whether an AI system has genuinely solved a biological problem or arrived at an answer through superficial pattern matching.
• Review benchmark questions for ambiguity, unintended shortcuts, insufficient information, or multiple valid interpretations.
• Provide structured feedback to improve benchmark quality and model evaluation.
• Use relevant scientific literature and established research methods to validate benchmark content.
• Contribute to benchmark development across relevant biological and biomedical domains.
• Work independently while following detailed quality, confidentiality, and evaluation guidelines.

Relevant Research Areas

Candidates may come from one or more of the following areas:

• Molecular Biology
• Cell Biology
• Genetics
• Genomics
• Proteomics
• Microbiology
• Developmental Biology
• Evolutionary Biology
• Ecology
• Physiology
• Biotechnology
• Bioinformatics
• Computational Biology
• Systems Biology
• Structural Biology
• Synthetic Biology
• Biomedical Sciences
• Cancer Biology
• Pharmacology
• Other closely related life-science disciplines

Candidates with deep expertise in a specific research area are encouraged to apply rather than requiring broad knowledge across every biological discipline.

What Makes a Strong BioMystery Problem?

Strong benchmark problems should test scientific reasoning rather than simple recall.

For example, a strong problem may require a researcher to:

• Connect findings across multiple biological mechanisms.
• Compare competing hypotheses.
• Identify the most plausible explanation from incomplete evidence.
• Reason from experimental results to an underlying mechanism.
• Detect an experimental design flaw.
• Predict the consequences of a biological intervention.
• Integrate information from different biological domains.
• Distinguish correlation from causation.
• Identify an inconsistency in an apparently plausible scientific explanation.
• The objective is to identify situations where an AI system must reason like a scientific researcher, rather than simply retrieve a memorized fact.

Qualification path

Mandatory Qualifications

• PhD completed or currently pursuing a PhD in Biology, Biotechnology, Biochemistry, Biomedical Sciences, Life Sciences, or a closely related discipline.
• Demonstrated research experience in a relevant biological or biomedical field.
• Strong understanding of advanced biological concepts within the candidate's area of expertise.
• Ability to critically evaluate scientific evidence and research conclusions.
• Strong analytical and scientific reasoning skills.
• Ability to identify subtle factual, conceptual, methodological, and logical errors.
• Excellent written and spoken English.
• Ability to explain complex biological concepts clearly and precisely.
• Strong attention to detail.
• Ability to work independently in a remote, project-based environment.

Research Evidence

Preference will be given to candidates with evidence such as:

• Peer-reviewed publications
• PhD research
• Research projects
• Conference presentations
• Laboratory research
• Scientific collaborations

Publication history is preferred but should not be an automatic requirement if the candidate can demonstrate substantial doctoral-level research expertise.

Preferred Qualifications

• Postdoctoral research experience.
• Multiple peer-reviewed publications.
• Experience reviewing scientific literature or manuscripts.
• Experience with experimental design and data interpretation.
• Experience creating scientific assessments, benchmark questions, or research datasets.
• Experience with AI, LLM evaluation, AI-for-Science, or scientific benchmarking.
• Familiarity with computational biology or bioinformatics tools.
• Experience evaluating scientific reasoning or research outputs produced by AI systems.

Prior AI experience is not mandatory. Strong domain expertise and research reasoning are more important.

English proficiency is mandatory.

Candidates must be able to:

• Read and interpret scientific literature in English.
• Understand complex scientific questions.
• Explain biological reasoning clearly in English.
• Produce high-quality written benchmark content in English.
• Communicate effectively during the qualification and interview process.