Elite Research Scientist - Frontier AI Evaluation

The Opportunity

• We are assembling a closed, invitation-level network of exceptional scientific minds to contribute to frontier AI systems at the highest level of rigor.
• This is not annotation. This is scientific adjudication.
• We are selecting 10 individuals per discipline in:
• Mathematics
• Physics
• Neuroscience
• Chemistry
• Admission is limited. Standards are uncompromising.

The Profile

• We are looking for individuals whose credentials place them in the top percentile of their field.

Examples of the caliber we seek

• International Olympiad medalists such as IMO, IOI, IPhO, IChO or equivalent.
• Professors or faculty at top-ranked universities.
• PhD holders from globally recognized institutions.
• Researchers with peer-reviewed publications and strong citation impact.
• Scientists with demonstrated original contributions to their field.
• Individuals recognized for exceptional analytical depth and intellectual rigor.
• We are particularly focused on identifying outstanding researchers in developing regions whose academic accomplishments rival leading global institutions.
• This is a curated network. We are not scaling volume. We are selecting excellence.

The Work

• Selected researchers will:
• Evaluate advanced scientific reasoning produced by state-of-the-art AI systems
• Assess multi-step derivations, proofs, theoretical arguments, and domain-specific outputs
• Identify subtle conceptual flaws or logical inconsistencies
• Provide research-grade corrective feedback
• This requires deep subject mastery. Surface familiarity is insufficient. The standard is publication-level rigor.

Why this is Distinct

• Direct exposure to frontier AI research challenges
• Collaboration with an elite peer group across disciplines
• Application of scientific expertise beyond academia
• Flexible, intellectually demanding engagement
• Compensation aligned with top-tier expertise
• This is not transactional work. It is high-level intellectual contribution.

Candidates must demonstrate the following

• PhD or near completion in Mathematics, Physics, Neuroscience, or Chemistry
• Peer-reviewed publications in respected journals
• Meaningful citation record indicating research impact
• International academic distinction such as Olympiad medalist
• Faculty position at a leading university

All credentials will be verified. A technical screening will assess depth of reasoning. Selection is competitive and limited to a small cohort per field.

Our Positioning

We are building a network of the most accomplished scientific researchers in developing regions, individuals whose academic records match global standards and whose intellectual drive makes them exceptional contributors to frontier AI systems.

If you believe your credentials and analytical rigor place you among the strongest researchers in your field, we invite you to apply.

Application Submission

If you believe your credentials and analytical rigor place you among the strongest researchers in your field, we invite you to submit your formal application.

Please complete the required data form to begin the confidential review process: https://forms.gle/4zF93iDXnSRN35YL7 .

About Perle

Perle powers the most ambitious AI initiatives in the world with human intelligence at scale. We work with the world’s leading model builders and enterprises to deliver expert-in-the-loop data, model evaluation, and trust systems that make AI safe, responsible, efficient and exceptionally high-performing.

Our clients don’t buy software, they buy outcomes: accuracy, reliability, and defensibility.

We’re an early-stage, founder-led company operating in the highest-stakes markets of AI. Every hire defines how we win.

Where you can work

This role is 100% remote, for applicants located in these 24 countries: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, Egypt, Guatemala, Honduras, India, Mexico, Nicaragua, Panama, Peru, Philippines, Paraguay, El Salvador, Thailand, Türkiye, Uruguay, Venezuela and Vietnam.