Quality assurance, multilingual labeling (EN–ID)
- Labeled and categorized multimodal datasets (text, images, audio) based on detailed project guidelines. - Performed object detection, entity recognition, and text classification, ensuring accuracy and consistency across large volumes of data. - Converted unstructured data into structured formats (numerical/text vectors, image bounding boxes) for AI/ML training. - Conducted quality assurance checks to eliminate errors, validate annotations, and maintain dataset reliability. - Designed and tested prompts in Bahasa Indonesia, assessing AI responses for accuracy, fluency, cultural relevance, and alignment with instructions. - Provided structured feedback to improve AI’s ability to handle Indonesian language tasks, including grammar, semantics, and contextual appropriateness. - Collaborated in iterative data refinement after model testing, supplying additional or corrected data to enhance model recognition and reduce bias.