Audio Model Trainer: Mercor

This role is part of a cutting-edge AI research initiative focused on building and refining datasets for multimodal AI systems. The successful candidate will be responsible for recording short audio clips that describe visual content, directly contributing to the development of next-generation models that can understand and interact with the world across both visual and auditory domains.

Responsibilities:

View a series of images and generate clear, concise, and natural-sounding spoken descriptions.

Record short, high-quality audio clips (typically 2-3 minutes each) that are free from background noise.

Follow specific linguistic, timing, or stylistic guidelines provided by the research team.

Collaborate with AI researchers and QA teams to review and iterate on data quality.

Requirements (Qualifications):

Excellent verbal communication and enunciation skills.

Native or near-native fluency in English (fluency in other languages is a plus).

Strong attention to detail and the ability to follow guidelines precisely.

Prior experience with voice recording or data annotation is a plus, but not required.

Comfortable working independently and handling repetitive tasks with consistency.