
AI Annotator
Deccan AI · Dec 2025 – Present
- Worked as an AI Annotator at Deccan AI, where I evaluated and compared responses generated by large language models (LLMs) to improve their performance and accuracy.
- My role involved assessing multiple model outputs based on defined evaluation rubrics such as groundedness, instruction following, and truthfulness.
- I analyzed whether responses were factually correct, relevant to the prompt, and aligned with user intent.
- I identified issues like hallucinations, incomplete answers, and deviations from instructions, and provided structured feedback to enhance model reliability.
- This process directly contributed to refining training datasets and improving the overall quality of AI systems.
- Through this experience, I developed strong analytical skills, attention to detail, and a deep understanding of how human feedback is used to train and optimize AI models.




