Responsibilities:

Prompt Validation & Data Annotation

• Execute Multi-Step GenAI Tasks: Accurate interpretation and systematic deployment of complex generative sequences (e.g., classify\(\rightarrow\) reason\(\rightarrow \) respond).
• Interpret Ambiguity: Decode loosely defined client guidelines or vague initial guidelines into clear, functional instructions for model alignment.
• Labeling Taxonomy Application: Categorize complex data streams systematically, establishing pattern recognition across massive datasets.

AI Reasoning, Safety & Quality Assurance

• Detect Model Vulnerabilities: Identify edge cases, logical inconsistencies, and factual hallucinations within model outputs.
• Rubric-Based Evaluation: Grade and critique complex AI outputs against detailed qualitative rubrics for accuracy, relevance, and safety.
• Fact-Checking & Synthesis: Perform deep primary research to validate data claims, compressing findings into highly structured and usable training datasets.

Iterative Refinement & Collaboration

• Probing & Optimization: Engineer logical follow-up prompts and iterative inquiries to test the boundaries of and maximize model capability.
• Instructional Communication: Translate highly technical or abstract model behaviors into structured, actionable feedback loops for development pipelines.
• Human-in-the-Loop Pipeline: Collaborate closely with engineering and shared data teams to implement policy and model updates.

Qualifications:

Linguistic & Writing Command

• Advanced Writing Proficiency: C1-level written capability or equivalent mastery, displaying clean text syntax, structural accuracy, and stylistic adaptability.
• Creative & Structured Text Design: Experience drafting diverse copy styles, ranging from imaginative narrative tasks to highly rigid, formatted technical instructions.

Cognitive & Analytical Abilities

• High Reading Comprehension: Exceptional capability to ingest and breakdown dense, sophisticated text data quickly.
• Nuanced Decision Making: High comfort level operating within ambiguous contexts, using partial data or conflicting guidelines to form sound, objective conclusions.
• Critical Evaluation: Innate analytical skepticism needed to separate biased or unsafe outputs from objective text.

GenAI Behavioral & Cognitive Readiness

• Cognitive Resilience: Proven ability to maintain speed and precision when processing high-volume data streams under demanding cognitive load.
• Learning Agility: Proactive curiosity regarding AI technologies, with a natural inclination to adapt swiftly to changing model updates, rules, and ecosystem dynamics.
• Ethical Judgment & Bias Awareness: Strict commitment to personal integrity, maintaining high emotional intelligence to detect structural or societal bias within AI behavior.