Juliet Eleazu Lagos

Reviewed daily LLM outputs across diverse topics, scoring for accuracy, helpfulness, and safety. Tested different prompts to identify failure points, off-topic responses, and hallucinations. Provided clear, specific feedback to developers explaining why an output failed and how to improve it. Verified facts against reliable sources to ensure responses were grounded in truth and met quality expectations.

Evaluate AI-generated English and Igbo responses for accuracy, fluency, relevance, grammar, and cultural appropriateness. Review chatbot conversations and voice-model outputs for natural language quality. Annotate and label multilingual language data according to project guidelines, identify errors in translation/intent/context/terminology, and provide structured written feedback to improve AI model performance. Conduct audio and language quality assessments and support multilingual dataset quality.

Performed RLHF evaluation tasks by comparing two or more AI options and selecting the best response based on user intent. Flagged harmful, biased, or inappropriate outputs following strict moderation guidelines. Rated search query results and categorized text passages for training NLP models. Maintained high accuracy during weekly quality audits while meeting tight task deadlines.