Chemistry Expert - AI Trainer (Chile)

📋 Description

• • Design advanced, deterministic chemistry problems with exactly one verifiable correct answer for training and evaluating frontier AI models.
• • Create reasoning-intensive tasks grounded in real scientific workflows across computational chemistry, physical chemistry, analytical chemistry, and chemical engineering-adjacent domains.
• • Develop technically precise problems that require deep domain knowledge, numerical reasoning, and computational rigor beyond textbook recall.
• • Write clear, concise, and rigorous technical explanations in English for all problem sets, including step-by-step solutions and validation criteria.
• • Use Python to implement problem logic and where appropriate, integrate domain-specific chemistry tools or simulation packages such as RDKit, Gaussian, ORCA, or similar.
• • Ensure all generated tasks are fully reproducible, with documented inputs, parameters, and output expectations for consistent AI evaluation.
• • Collaborate with AI training teams to align problem complexity and structure with model evaluation needs.
• • Maintain high standards of technical accuracy and consistency across all deliverables to support high-stakes AI benchmarking.
• • Document problem design decisions, assumptions, and validation methods to enable auditability and scalability of the training dataset.
• • Deliver high-quality content on a strict schedule within a 2-month contract period, with potential for extension based on performance.
• • Work independently with minimal supervision while adhering to strict quality control protocols for scientific integrity.
• • Adapt problem designs based on feedback from AI model performance and evaluation metrics provided by the team.
• • Maintain confidentiality of proprietary training data and evaluation frameworks used by leading AI labs.
• • Communicate clearly and promptly regarding project progress, challenges, and deliverables in a remote, asynchronous environment.
• • Ensure all content is free from ambiguity, bias, or non-deterministic outcomes that could compromise AI training reliability.
• • Apply expertise in chemical reasoning to construct novel problems that test AI systems on synthesis planning, reaction prediction, spectral interpretation, thermodynamic analysis, or molecular property calculation.
• • Verify solutions using established computational methods and cross-check results against authoritative chemical databases or literature.
• • Contribute to the continuous improvement of problem design standards by identifying gaps in current AI capabilities and proposing targeted challenges.