Chemistry Research AI Evaluator

About OpenTrain

OpenTrain is the #1 platform for building careers in AI training and data labeling. We help specialists find and manage AI-training work, consolidate proof-of-work across projects, and grow a durable freelance portfolio. Creating an OpenTrain account is free.

About AI training in chemistry

AI training is the human side of building modern models: experts teach models by reviewing outputs, authoring examples, and grading scientific reasoning. In chemistry-focused work you influence how models perform on retrosynthesis, mechanism prediction, spectral interpretation, computational chemistry, and literature reasoning.

This work is a cutting-edge way to apply research expertise outside the lab: most projects are remote and flexible, and contributors directly shape how state-of-the-art scientific AI systems behave.

The role

OpenTrain is hiring Chemistry Research AI Evaluators to perform high-skill evaluation and content-authoring tasks for frontier scientific reasoning work. This is a part-time contractor role requiring 20+ hours per week, available to applicants located in the United States. Pay is hourly at $80–$110 USD, depending on experience.

Contributors collaborate with a small program team and other senior specialists on tasks that require active research judgment and publication-level expertise.

• Employment type: Contractor, Part-time
• Location: Remote within the US (must be based in the United States)
• Time commitment: 20+ hours/week
• Pay: $80–$110 USD per hour

What you'll do

• Review chemistry research papers for correctness, novelty, and significance.
• Author and review challenging chemistry problems for AI training and evaluation.
• Evaluate model outputs on retrosynthesis, mechanism prediction, spectral interpretation, computational chemistry, and literature reasoning.
• Identify failure modes and errors that a practicing researcher would catch.
• Collaborate with a small program team and other senior contributors to refine tasks and grading rubrics.

Requirements

• First-author or co–first-author publication record in respected chemistry journals.
• Doctorate (PhD) or postdoctoral training in chemistry or a related STEM field.
• Current or recent active research experience as a PhD candidate, postdoctoral researcher, instructor, faculty member, principal scientist, or industry researcher.
• Ability to judge correctness, novelty, and significance in research papers.
• Experience evaluating model outputs or scientific reasoning tasks.
• Deep background in one or more of organic synthesis, catalysis, physical chemistry, computational chemistry, or materials chemistry.

Helpful background and examples

• Publications in journals such as JACS, Nature Chemistry, Science, Nature, Angewandte Chemie, Nature Communications, ACS Catalysis, Chem, Chemical Reviews, Chem Soc Rev, JOC, Organic Letters, or Inorganic Chemistry.
• Experience with patents, grants, competitive fellowships, or international chemistry olympiad medals is a strong plus.
• Hands-on experience in organic synthesis, catalysis, structural chemistry, chemical biology, theoretical/computational chemistry, or materials chemistry.

How work is structured

This project focuses on text-based chemistry evaluation and authoring. Label types include evaluation ratings and text-generation tasks; data you'll work with is primarily research text and model outputs.

You will be contracted and paid hourly. Contributors typically interact with a small program team, follow written rubrics, and participate in reviewer calibration to ensure consistent, high-quality judgments.

• Data type: Text
• Labeling tasks: EVALUATION_RATING and TEXT_GENERATION
• Language: English

Who should apply and next steps

Apply if you are an active chemistry researcher with a strong publication record, doctoral or postdoctoral training, and experience judging scientific work. Outstanding senior PhD students and recent graduates who meet the publication and research criteria are encouraged to apply.

To get started, create or update your OpenTrain profile, include your publication record and relevant research experience, and indicate availability for a 20+ hour weekly commitment. OpenTrain helps you centralize proof-of-work so your contributions can build a visible, long-term AI-training portfolio.