AI Trainee – AI Evaluation, Annotation and Quality Assurance
As an AI Trainee at Deccan AI Private Ltd, I participated in various data labeling and evaluation projects focused on large language and agent-based systems. My tasks involved reviewing AI outputs, annotating prompt responses, benchmarking performance, and performing multi-step quality assurance. I evaluated agent decision logic, stress-tested scenario instructions, and ensured annotation consistency and reliability. • Conducted structured annotation and evaluation of AI-generated text and agent interactions. • Identified edge cases, ambiguities, and failure modes in scenario designs. • Documented findings and provided quality feedback for iterative improvement. • Gained direct experience with LLM evaluation workflows and structured quality checks.