About the role
We are seeking detail-oriented and analytical freelancers to support the training and evaluation of Large Language Models (LLMs) by reviewing, annotating, and evaluating video and audio content. In this role, you will analyze multimedia data, assess descriptive details, and provide high-quality annotations that help improve AI models' understanding of visual and audio information. The ideal candidate has excellent English language skills, strong observational abilities, and the capability to accurately evaluate content while adhering to detailed project guidelines.
Key Responsibilities Review and annotate video and audio content according to project guidelines. Evaluate the accuracy, completeness, and quality of AI-generated descriptions for multimedia content. Compare multiple AI-generated responses and determine the most accurate and contextually appropriate output. Identify key visual and audio elements, including objects, people, actions, events, environments, emotions, sounds, and contextual details. Assess descriptive accuracy, relevance, consistency, and completeness of annotations. Label and categorize multimedia content using designated annotation tools. Provide objective feedback to improve AI-generated outputs. Follow annotation guidelines and quality standards while maintaining consistency across tasks.
Bachelor's degree in any courseExcellent English communication, reading, and comprehension skills.
Excellent attention to detail and observational skills.Ability to interpret visual and audio information accurately. Comfortable working with structured guidelines and quality standards. Reliable internet connection and personal computer laptop. Benefits: Flexible schedule Work Location: Remote
From the employer's public posting. AI Eval HQ isn't affiliated with Innodata; you apply on their site.