The company is looking to solve the problem of creating and refining specialized training data for advanced AI models in the field of physics, specifically in Classical Mechanics, Quantum Mechanics, and Thermodynamics.
Requirements
- Strong understanding of Classical Mechanics, Quantum Mechanics, and Statistical/Thermodynamic systems.
- Familiarity with AI/ML concepts, or prior experience in AI model training or evaluation.
- Proficiency in Python, LaTeX, or scientific computing tools such as Mathematica or Jupyter.
- Experience with data annotation, prompt engineering, or language model fine-tuning.
Responsibilities
- Create, annotate, and evaluate complex training data sets in physics across multiple modalities, including text, images, and video.
- Act as a subject-matter expert in Classical Mechanics, Quantum Mechanics, and Statistical/Thermodynamic systems, guiding AI model development and learning.
- Craft clear, precise, and pedagogically sound explanations and problem solutions in diverse academic and instructional styles.
- Audit and improve AI-generated content, ensuring adherence to rigorous scientific and linguistic standards.
- Perform quality checks on tasks completed by other experts, ensuring consistency and accuracy in physics content.
- Collaborate closely with machine learning scientists and engineers to enhance data guidelines, identify edge cases, and optimize annotation tools.
- Troubleshoot data or guideline ambiguities, contributing high-judgment feedback to continuous process improvement efforts.
Other
- Master's or PhD in Physics or closely related field.
- Strong proficiency in English. Candidate must demonstrate excellent writing, reading, and comprehension skills (C2 level in the Common European Framework CEFR scale).
- Strong understanding of U.S.-based culture, society, and norms.
- Exceptional written and verbal communication skills, capable of explaining complex concepts clearly and concisely.
- Demonstrated ability to work independently, exercising sound judgment in ambiguous or incomplete scenarios.
