AI Content Annotator (Tier 1 & Tier 2) | Outlier AI
As an AI Content Annotator at Outlier AI, I specialized in reinforcing learning from human feedback (RLHF) and supervised fine-tuning (SFT) to improve the accuracy and safety of large language models (LLMs). I leveraged over 20 years of business and financial expertise to evaluate and rank AI-generated responses in the domains of business strategy, asset management, and financial forecasting. My prompt engineering and evaluation responsibilities included crafting complex prompts and providing detailed, rubric-based justifications for model rankings, with a focus on truthfulness and logical coherence. • Conducted multi-tier annotation tasks and progressed rapidly from Tier 1 to Tier 2 due to high quality scores. • Evaluated LLM outputs, ensuring alignment with professional standards for safety and accuracy. • Specialized in business and financial subject matter, applying domain expertise to data labeling work. • Employed prompt engineering to test model boundaries and enhance model performance.