[0619PJXS] Agentic Life Science Expert

ZettaMine Labs Pvt. Ltd.; India; Remote; INR 1,674,000 - 2,344,000; Full time

Job summary

ZettaMine Labs Pvt. Ltd. invites applications for Agentic Life Science Expert on a short-term AI training and evaluation project across India. The role is remote, contractor-based, with immediate start and PST overlap requirements.

You will design authentic multi-step life science tasks, translate workflows into reproducible environments, and develop robust benchmarks. Experience with Python, Linux, bioinformatics, and computational biology is essential.

Qualifications

• Strong programming skills in Python with command-line tools and scientific libraries.
• Experience in Linux or terminal-based environments.
• Deep knowledge of computational life sciences and multi-step analyses.
• Ability to implement, test, debug, and validate computational workflows.
• Ability to validate scientific reasoning, calculations, code, intermediate results, and final outputs.
• Develop rigorous scientific tasks with defined inputs, outputs, constraints, and ground truths.
• Experience with scientific datasets, biological data, computational models, and domain-specific tools.
• Ability to develop multi-step workflows in bioinformatics, genomics, systems biology, neuroscience, biostatistics, and related fields.
• Troubleshoot intermediate results, dependencies, data-processing issues, and output inconsistencies.
• Develop reproducible and verifiable scientific workflows.

Responsibilities

• Design authentic, multi-step Agentic Life Sciences tasks based on realistic workflows.
• Translate scientific workflows into self-contained computational environments.
• Create challenging scientific workflows across life sciences requiring multiple steps.
• Prepare realistic input files, datasets, instructions, constraints, resources, and deliverables.
• Implement expert solutions using Python, CLI tools, libraries, and domain software.
• Create tasks involving bioinformatics, genomics, systems biology, computational neuroscience, and related fields.
• Build tasks requiring AI agents to inspect data, select methods, execute analyses, troubleshoot problems, and synthesize outputs.
• Develop reproducible expert solutions and objective ground truths.
• Develop robust grading criteria for scientific outputs and ensure evaluation quality.
• Ensure tasks are self-contained, reproducible, and executable in network-isolated environments.

Skills

Python

Linux

Bioinformatics

Multi-step Analyses

Reproducible Workflows

Debug & Validate

Docker

AI Systems

Datasets Handling

Education

Ph.D. in Life Sciences

Postdoctoral experience

Advanced research experience

Tools

Python libraries

Command-line tools

Scientific computing

Job description

We are hiring Agentic Life Science Experts for a short-term AI Training & Evaluation project across: India

Role: Agentic Life Science Expert

Experience: Ph.D., Postdoctoral Experience, or Equivalent Advanced Research Experience

Mode: Remote

Engagement: Contractor / Short-Term Contract

Start: Immediate

Mandatory Eligibility Criteria

Technical Qualification

• Strong scientific programming skills in Python, command-line tools, scientific libraries, or other relevant computational technologies.
• Experience working in Linux or terminal-based environments.
• Strong knowledge of computational life sciences, scientific computing, computational research methodologies, and multi-step scientific analysis.
• Ability to independently implement, test, debug, and validate computational scientific workflows.
• Ability to independently validate scientific reasoning, calculations, code, intermediate results, and final computational outputs.
• Ability to develop rigorous scientific tasks with clearly defined inputs, expected outputs, constraints, ground truths, and evaluation criteria.
• Experience working with scientific datasets, biological data, computational models, input files, research pipelines, and domain-specific scientific tools.
• Ability to develop and validate multi-step computational workflows involving bioinformatics, genomics, systems biology, neuroscience, biostatistics, drug discovery, biochemistry, structural biology, protein engineering, or microbiology.
• Ability to troubleshoot intermediate results, dependencies, scientific assumptions, computational errors, data-processing issues, and output inconsistencies.
• Ability to develop reproducible and objectively verifiable scientific workflows.
• Ability to design tasks that are self-contained and executable in controlled, network-isolated computational environments.
• Ability to validate scientific outputs for accuracy, consistency, reproducibility, and scientific correctness.

Educational Qualification

Mandatory: Ph.D., postdoctoral experience, or equivalent advanced research experience in Life Sciences, with strong demonstrated computational and scientific programming experience.

Preferred: Advanced research or professional experience in Bioinformatics, Computational Genomics, Systems Biology, Computational Neuroscience, Biostatistics, Computational Drug Discovery, Computational Biochemistry, Structural Biology, Protein Engineering, Computational Microbiology, Computational Life Sciences, or related fields.

Availability

• Minimum 4 hours of PST overlap per day
• Ability to work remotely on a short-term AI training and evaluation project.
• Immediate availability preferred.
• Ability to collaborate with project reviewers and incorporate feedback.
• Availability according to project requirements and deadlines.

Others

• Personal laptop/desktop and stable high-speed internet.
• Strong written and verbal communication skills.
• Strong scientific programming experience, particularly in Python.
• Experience working with scientific libraries, command-line tools, Linux, and computational research environments.
• Experience performing multi-step computational scientific analyses.
• Experience developing reproducible scientific and research pipelines.
• Ability to ensure computational tasks are self-contained, reproducible, and executable in controlled network-isolated environments.
• Ability to package required dependencies, scientific data, and computational resources within the project environment.
• Strong attention to scientific accuracy, documentation, reproducibility, and computational validation.
• Experience with Docker/Linux environments is preferred.
• Experience with AI agents, coding agents, or scientific AI systems is advantageous.
• Experience building automated evaluation environments, scientific benchmarks, or automated graders is preferred.
• Publications involving computational or data-intensive life sciences research are advantageous.
• Ability to translate authentic research workflows into bounded and objectively gradable scientific tasks.

What You’ll Work On

• Design authentic, multi-step Agentic Life Sciences tasks based on realistic research and computational workflows.
• Translate scientific and life sciences research workflows into self-contained computational environments.
• Create challenging and realistic scientific workflows across the life sciences that require multiple computational steps rather than simple question-and-answer or single-script solutions.
• Prepare realistic input files, scientific datasets, instructions, constraints, computational resources, and expected deliverables.
• Implement expert solutions using Python, command-line tools, scientific libraries, and relevant domain-specific software.
• Create tasks involving bioinformatics, computational genomics, systems biology, computational neuroscience, biostatistics, computational drug discovery, computational biochemistry, structural biology, protein engineering, and computational microbiology.
• Build tasks requiring AI agents to inspect scientific data, select appropriate methods, execute analyses, troubleshoot problems, interpret intermediate results, and synthesize final outputs.
• Develop reproducible expert solutions and objectively verifiable ground truths.
• Develop robust automated or semi-automated grading criteria for scientific outputs.
• Ensure tasks evaluate scientific reasoning and execution rather than memorization.
• Validate scientific assumptions, calculations, code, intermediate outputs, and final answers.
• Validate that tasks are self-contained, reproducible, and execute successfully in controlled, network-isolated computational environments.
• Document scientific assumptions, computational requirements, dependencies, expected outputs, evaluation criteria, and known limitations.
• Maintain high quality and throughput while incorporating feedback from project reviewers.
• Develop benchmark tasks that evaluate whether AI agents can independently navigate scientific data, reason through complex life sciences problems, write and execute code, use computational tools, troubleshoot intermediate results, perform multi-step scientific analyses, and produce accurate, reproducible, and objectively verifiable scientific outputs.