Subject Matter Expert – Quantative/Scientific/Corporate Domains – Marathi Remote

Role overview

Create and review realistic technical, scientific, and corporate work scenarios in Marathi for an AI benchmarking project. The work is grounded in local or regional operating contexts and tests whether responses reflect professional practice rather than generic recall. Tasks combine domain judgment with language-specific context across software engineering, mathematics, robotics, healthcare, natural science, manufacturing, finance, media, and office work. Contributors design the work, adapt scoring criteria, review responses, and produce defensible reference solutions.

What you will deliver

• Design realistic scenarios in Marathi or English.
• Adapt structured evaluation rubrics.
• Review AI-generated and human responses for factual accuracy, quality, cultural appropriateness, and operational realism.
• Develop reference solutions reflecting local and domain practice.
• Understand, refactor, and verify code execution in a sandboxed environment when the task requires it.

Domain coverage

• Software engineering and mathematics.
• Robotics engineering.
• Healthcare and natural science.
• Manufacturing and industrial systems.
• Finance and financial analysis.
• Media and content.
• Office and corporate work.

Evaluation standards

• Factual accuracy.
• Domain realism.
• Cultural appropriateness.
• Technical execution when required.
• Consistent rubric use.

Qualifications

• Native-level Marathi with strong professional English proficiency.
• At least five years in the relevant technical, scientific, and corporate work.
• Localized fluency and strong professional English.
• Deep contextual knowledge of the chosen domain and its workflows.

Contract terms

• Independent-contractor engagement.
• Contractors may choose when and how much to work; no fixed hours, check-ins, or guaranteed volume are stated.