Scientific Coding AI Training Expert

About OpenTrain

OpenTrain AI is the hiring and contracting organization for this role. It helps people build careers in AI training and data labeling by connecting specialized contributors with projects, supporting professional profiles, and making it easier to grow a durable portfolio of AI work.

About AI Training Work

AI training is the human side of building modern artificial intelligence. Contributors create examples, review model outputs, and evaluate whether generated content or code is correct, useful, and reliable. Scientific coding experts help advanced models improve by providing rigorous problems, validated solutions, and high-quality assessments.

• Work on cutting-edge AI development through human-led task creation and model evaluation.
• Use specialized STEM knowledge to assess correctness, clarity, determinism, and scientific validity.
• Remote, flexible AI training work can help you build experience in a fast-growing technology field.

The Role

OpenTrain is seeking a Scientific Coding AI Training Expert to create and evaluate complex scientific programming tasks used to train and assess advanced AI models. You will turn challenging STEM problems into structured specifications, implement verified Python solutions, develop tests that distinguish strong and weak model outputs, and refine work against demanding quality standards.

This is a freelance contractor assignment focused on scientific coding task creation and evaluation. The engagement is a focused short-term project requiring 20+ hours per week.

• Experience level: Intermediate
• Work type: Freelance contractor and part-time
• Schedule: 20+ hours per week
• Working language: English
• Eligible locations: Bangladesh, Brazil, Colombia, Egypt, Ghana, India, Pakistan, Indonesia, Kenya, Nigeria, Türkiye, and Vietnam

What You'll Do

You will create technically rigorous programming content and assess the quality of AI-generated solutions. The work combines scientific writing, Python implementation, test design, structured review, and collaboration with project teams.

• Write scientific problem specifications with one main problem and at least three logically connected sub-problems that build toward the overall solution.
• Implement verified Python solutions with complete unit test coverage.
• Design discriminative test cases that distinguish correct from incorrect model outputs.
• Assess scientific validity, determinism, and expected results in coding tasks and generated solutions.
• Run structural and quality checks against defined rubrics, then revise tasks based on quality feedback.
• Evaluate model-generated outputs for correctness, clarity, and reliability.
• Participate in review discussions, feedback sessions, and project syncs.

Required Qualifications

This role requires an advanced academic and technical background. You should be able to formulate difficult scientific problems precisely, implement dependable solutions, and judge whether code and model outputs meet rigorous scientific and software-quality standards.

• Master's degree or PhD in a STEM discipline.
• Strong Python programming and scientific computing skills.
• Experience with NumPy, SciPy, SymPy, or comparable scientific tools.
• Ability to write rigorous scientific problems with clear constraints and expected outputs.
• Ability to create connected sub-problems that build toward an overall solution.
• Experience designing unit tests and discriminative cases that separate correct from incorrect code.
• Careful attention to scientific correctness, test discriminativeness, determinism, and task quality.
• Experience in AI data annotation, scientific research, or scientific writing.
• Familiarity with LLM evaluation frameworks or coding benchmarks.
• Published research or academic project experience in a STEM field.

Helpful Background

Experience creating evaluation datasets, reviewing model-generated code, or working with scientific programming libraries is useful. Background in biology or another specialized STEM field is also relevant when developing domain-specific problems.

• Scientific research or academic project work in a STEM field.
• Evaluation dataset creation.
• Review of model-generated code.
• Scientific programming library experience.
• Biology or other specialized STEM knowledge.

Why Work With OpenTrain

OpenTrain gives AI training professionals a place to develop and manage a credible record of specialized work. By building your profile and contributing to projects like this one, you can showcase your expertise, discover opportunities aligned with your skills, and grow toward a lasting career in AI training and data labeling. Creating an OpenTrain account is free.

• Build a portfolio around advanced scientific coding and AI evaluation work.
• Showcase specialized experience instead of starting from scratch for every opportunity.
• Contribute directly to how state-of-the-art AI systems learn and perform.
• Apply in minutes through OpenTrain.