Apply on Employer Site

Innodata Inc. · 1 week ago

Generative AI Specialist - Red Team (English and French)

Canada

Part-time

Remote

New Grad, Entry Level

CAD$25/hr - CAD$25/hr

Innodata Inc. is a leading data engineering company specializing in AI technology solutions. They are seeking a Generative AI Specialist to evaluate and enhance the performance of AI models, contribute to cutting-edge AI development, and assist in the training of large language models while enjoying a flexible, part-time work schedule.

AnalyticsArtificial Intelligence (AI)Enterprise SoftwareInformation TechnologyInternetMachine Learning

Responsibilities

Evaluation: Rating/assessing the performance of AI models or algorithms based on their output or behavior through a set of evaluative questions

Annotation Labeling: Labeling elements of a piece of content rather than the content as a whole

Classification: Assigning predefined categories or labels to items

Content Quality: Evaluating the perceived quality and/or appropriateness of content

Content Understanding: Generating labels to advance understanding of a concept, trend etc

Data Augmentation: Creation of additional training data for machine learning models by applying transformations to the original data, such as modifying images (rotation, flipping, cropping), generating new text (paraphrasing, summarization), or altering audio/video signals (speed modification, pitch shifting) to reduce overfitting and increase dataset diversity

Grading: Reviewing data and identifying whether or not a product feature works as intended based on the project's guidelines

Identification Labeling: Labeling model outputs to identify if a piece of content is or isn't something. Examples: identify clickbait; identifying gaming videos; identifying branded content

Preference Ranking: Ordering or ranking items based on a set of preferences or criteria

Prompt Generation: Creating prompts or questions that will be used to generate responses from a language model or other AI system

Relevance Evaluation: Projects that evaluate the relevance of content based on a relevancy scale (1-3, 1-5, etc.)

Response Generation: Generating responses to prompts or questions using a language model or other AI system

Response Rewrite: Rewriting existing text while preserving the original meaning, often to improve clarity or style and adherence to guidelines

Response Summarization: Producing concise summaries of longer pieces of text or data

Similarity Evaluation: Projects where content is compared in order to drive a determination

Transcription: Converting spoken language or audio content into written text

Translation: Converting text or spoken language from one language to another

Data Collection: Gathering and compiling various forms of data to be used for training, evaluating, or fine-tuning the AI models. This may include text, images, videos, audio files, or other types of digital content

Qualification

Generative AILarge Language ModelsData AugmentationHumanities DegreeEnglish ProficiencyFrench ProficiencyContent EvaluationAnnotation LabelingResponse GenerationTranslation