Remote Generative AI Specialist (Part-Time, Project-Based)

Innodata; United States; Hybrid; USD 34,000 - 41,000; Part time

Job summary

Innodata is seeking a Generative AI Specialist to join its on-demand SME program. This part-time, remote role lets you contribute to advanced AI development on your schedule, collaborating with global teams and helping LLMs learn language and reasoning.

You will evaluate models, label data, create training data, and craft prompts, all while maintaining high-quality, scalable outputs from anywhere in the world.

Qualifications

• PhD is required.
• Proficiency in English at expert level (C1/C2).

Responsibilities

• Rating or evaluating AI model outputs based on provided guidelines.
• Labeling content elements rather than the whole content.
• Assigning predefined categories or labels to items.
• Evaluating the quality and appropriateness of content.
• Generating labels to advance understanding of concepts or trends.
• Creating additional training data via transformations (text, images, audio/video).
• Reviewing data to ensure alignment with project guidelines.
• Labeling model outputs to identify clickbait or branded content.
• Ordering or ranking items per predefined criteria.
• Creating prompts for language models used in responses.
• Evaluating content relevance on a scale.
• Generating responses to prompts using AI systems.
• Rewriting text to improve clarity while preserving meaning.
• Producing concise summaries of long text or data.
• Transcribing spoken language to written text.
• Translating text or speech between languages.
• Gathering diverse data types for training, evaluation, or fine-tuning.

Skills

English proficiency (C1/C2)

Education

PhD

Job description

Innodata is seeking a Generative AI Specialist to join its on-demand SME program. This part-time, remote role lets you contribute to advanced AI development on your schedule, collaborating with global teams and helping LLMs learn language and reasoning.

You will evaluate models, label data, create training data, and craft prompts, all while maintaining high-quality, scalable outputs from anywhere in the world.