Hindi Audio Evaluator — Transcription and Evaluation

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help people discover projects, build a unified AI-training portfolio, and grow a durable freelance career — and creating an OpenTrain account is free.

OpenTrain AI is hiring directly for this role. We hire contractors to work on annotation, evaluation, and QA projects that shape how modern AI systems behave.

Why AI training work matters

AI training (data labeling, annotation, and human evaluation) is the human side of building intelligent systems. Contributors transcribe audio, rate model outputs, create rubrics, and annotate language nuances that teach models to understand and generate better responses.

This work is remote-friendly and flexible, and it's an opportunity to directly influence how language models represent and use Hindi including dialects, code-switching, and regional expressions.

Role overview

We’re hiring a Hindi Audio Generalist Evaluator to transcribe Hindi audio and video, evaluate model responses, and help create rubrics and grading standards for consumer-facing language tasks.

This is a short-term, contractor engagement paying $50 per hour, roughly 10–20 hours per week, and requires candidates to be based in the San Francisco Bay Area for occasional in-person collaboration.

What you'll do

Work closely with OpenTrain project teams to produce high-quality transcriptions, evaluate model outputs against clear criteria, and prepare materials for benchmarking and QA.

• Transcribe Hindi audio and video following detailed constraints and style guidelines.
• Review and grade model responses for accuracy, fluency, completeness, and instruction-following.
• Develop and document evaluation rubrics and grading guidelines in Hindi and English.
• Support QA cycles, perform consistency checks, and prepare data for benchmark readiness.
• Capture tone, intent, register (formal/informal), regional expressions, dialect variation, and Hindi–English code-switching where relevant.

Requirements

You must meet the core language, location, and skill requirements listed below. We will preserve these requirements exactly as stated.

• Native or near-native fluency in Hindi (spoken and written).
• Professional fluency in English for written outputs and evaluation work.
• Experience transcribing or annotating Hindi audio or video (preferred).
• Ability to create clear evaluation rubrics and grading guidelines in Hindi and English.
• Familiarity with Hindi–English code-switching, regional dialects, and contemporary usage.
• Strong writing, editing, and critical thinking skills.
• Ability to work independently, manage time effectively, and meet deadlines.
• Based in the San Francisco Bay Area and available for occasional in-person collaboration.

Helpful background and who should apply

This position suits college students, recent graduates, and practitioners with backgrounds in linguistics, humanities, social sciences, journalism, translation/localization, or technical fields who want hands-on experience with language-model evaluation.

Interest in AI, practical exposure to evaluation workflows, or prior annotation experience will help you ramp up quickly, but strong Hindi skills and careful attention to linguistic nuance are the most important qualifications.

Pay, schedule, and logistics

Compensation and schedule are as follows; we will not deviate from the stated terms.

Contract type: Contractor, part-time. This is a short-term engagement.

• $50.00 per hour.
• Estimated 10–20 hours per week.
• Short-term engagement with occasional in-person meetings in the San Francisco Bay Area.

How it works

OpenTrain makes it easy to apply and manage AI-training work: create a profile, confirm your language and location, and apply. Successful applicants will receive project onboarding, style guides, and the evaluation templates you’ll use to complete tasks.

As a contractor with OpenTrain AI you will be paid at the stated hourly rate and expected to follow provided instructions and quality checks to ensure data and evaluation consistency.