Agentic AI Lead — Data Labeling and AI Training
As Agentic AI Lead at Swift Immersive, I worked daily with LLM outputs, gaining expertise in prompt engineering and understanding training data quality. My responsibilities included building RAG pipelines, document chunking, metadata tagging, and structured data ingestion, all closely related to annotation and data labeling. I also reviewed, validated, and corrected AI-generated responses to ensure consistency and accuracy in model outputs. • Developed and iteratively improved data labeling workflows for text and structured data • Performed prompt and response labeling for LLM instruction tuning and evaluation • Validated and quality-checked large volumes of structured data across spreadsheets and APIs • Provided feedback on model outputs to refine labeling guidance and improve system performance.