Deccan AI
|AI Systems Evaluator
remote, ,
Summary
Evaluated and refined AI models and datasets, contributing to the development of high-quality generative AI systems and improving instruction-following quality.
Highlights
Enhanced generative AI dataset quality by leveraging AI annotation and evaluation frameworks for supervised fine-tuning and preference optimization.
Improved Large Language Model alignment by applying Reinforcement Learning from Human Feedback (RLHF), evaluating and ranking hundreds of AI-generated responses for increased instruction-following and conversational consistency.
Streamlined modern LLM training workflows through collaborative iterative model feedback loops, encompassing prompt testing, comparative response analysis, and alignment scoring.