Join OpenTrain AI to define safety standards, escalation protocols, and abstraction frameworks for nuclear and radiological risk evaluation; remote, 20+ hrs/week, $50–$90/hr. Help shape how advanced models handle sensitive nuclear security information.
Generative AI & RLHF
100% Remote Hourly · $50–$90/hr
$50–$90/hr
Compensation
Worldwide
Eligibility
Entry
Experience
Jun 30, 2026
Posted
Open worldwide
Interested in this role?
Create a free OpenTrain account and apply in minutes.
OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help people start and grow freelance careers teaching AI by unifying opportunities, building a professional portfolio, and making it easy to apply and work across projects.
OpenTrain AI is the hiring and contracting organization for this role. We focus on meaningful, remote work that directly shapes how state-of-the-art AI systems behave.
About AI training work
AI training (data labeling, annotation, RLHF and evaluation) is the human side of building intelligent systems. Contributors annotate, evaluate, and provide feedback that models learn from, across images, audio, documents and model outputs.
This project centers on document evaluation and RLHF-style review: defining evaluation scope, handling sensitive technical information safely, and setting escalation and compliance rules for nuclear and radiological risks.
The role
We are recruiting a Nuclear & Radiological Security Expert to define standards for nuclear-safety evaluations, abstraction of sensitive information, escalation thresholds, and technical guidance that will be used to train and evaluate advanced models.
Work will focus on document-based evaluation and RLHF-style rating tasks: creating clear criteria, validating technical guidelines, and documenting decision rationales so models and human reviewers follow safe, compliant procedures.
Position type: Contractor, part-time.
Time commitment: 20+ hours per week.
Location: Remote; worldwide applicants accepted.
Data type: DOCUMENT; label types: EVALUATION_RATING and RLHF.
What you'll do
You will translate operational nuclear security expertise into concrete evaluation standards and workflows that human reviewers and models can follow. Your work will be documented and used to train evaluators and tune model behavior.
Define evaluation scope and standards for nuclear and radiological safety, nonproliferation, and security.
Build frameworks for abstracting or redacting sensitive/classified technical details in document-based settings.
Establish escalation thresholds and protocols for nuclear security risks in research and operational scenarios.
Validate and refine technical guidelines for nuclear and radiological safety and decision-making.
Document best practices and rationales for information security, compliance, and reviewer actions.
Collaborate with multidisciplinary stakeholders to align safety standards with national and global security interests.
Requirements
Candidates must demonstrate domain-specific knowledge and the ability to communicate technical guidance clearly for evaluators and model-training purposes.
Deep experience in nuclear physics, radiological safety, nonproliferation, or nuclear safeguards.
Background with government agencies, national laboratories, regulatory bodies, or academic research.
Understanding of nonproliferation frameworks, international safeguards, and weapons-relevant controls.
Experience with security evaluations, compliance protocols, or risk assessments in sensitive environments.
Strong technical writing and verbal communication skills; ability to document decision rationales.
Advanced degree, professional certification, or equivalent operational experience preferred.
Who should apply
This role is suitable for practitioners who can convert technical nuclear-security judgment into clear, repeatable evaluation rules for reviewers and models. Although listed as entry level, the role requires substantive domain knowledge as described above.
Early-career nuclear or radiological safety professionals who have worked in labs, regulators, or related research teams.
Subject-matter experts who want part-time, remote work contributing directly to AI safety and model evaluation.
People comfortable handling sensitive information under strict security and compliance expectations.
Compensation and time commitment
This contract role expects 20+ hours per week. Pay is hourly on a contractor basis with a stated range to reflect variability in task complexity and candidate background.
Pay: USD $50–$90 per hour (hourlyRate listed at $90; range shown reflects min–max).
Employment types: Contractor, Part-time.
Language requirement: English.
How the work will proceed
You will produce evaluation rubrics, secure abstraction/redaction frameworks, escalation rules, and documented examples. Tasks will be delivered as written standards, annotated examples, and reviewer guidance used in RLHF and evaluation workflows.
OpenTrain provides project access, task assignments, and payment processing; you will follow provided security and handling instructions for any sensitive content.
Join OpenTrain as a remote contractor to evaluate and red-team LLM outputs in French and English, focusing on safety, policy alignment, and adversarial case curation. This part-time role (20+ hrs/week) pays $24–$36/hr (typical $30/hr) and requires hands-on LLM red-teaming experience.
Join OpenTrain AI as a remote, part-time contractor reviewing and red-teaming LLM outputs in Hebrew and English to find safety failures and produce labeled evaluation data. $26–$38/hr, 20+ hours/week; your feedback will directly shape model safety.
Join OpenTrain to design and solve advanced physics problems that probe LLM reasoning and symbolic skills; remote, part-time contractor work (20+ hrs/week) for candidates with graduate-level physics experience and strong written English.
