AI Expert Annotator
Mindrift
May 2024 - Apr 2025 (1 year)
• Annotated and quality-assured RLHF datasets to train AI models for enhanced relevance, coherence, and adherence to ethical guidelines. • Designed, tested, and refined prompts for LLMs to improve response quality, ensuring outputs aligned with ethical guidelines and user intent. Applied iterative A/B testing to optimize prompt structures for tasks like legal reasoning, summarization, and creative generation. • Developed and refined annotation workflows and guidelines, ensuring consistency and scalability across projects. • Conducted iterative feedback cycles and analyzed annotation trends to identify data gaps and recommend improvements for greater model accuracy.