AI Safety Evaluator | Data Annotator | Linguistic QA & Rater | Market Research Specialist
Durban / Johannesburg, South Africa
Innovative and detail-driven AI Safety Evaluator and Multilingual Data Specialist with extensive experience in adversarial red teaming, content moderation, linguistic QA, and structured scoring. Skilled in LLM evaluation, audio annotation, prompt injection testing, and translation quality rating. Proven track record of improving AI model reliability, fairness, and user safety through human-in-the-loop evaluations, consumer persona testing, and multilingual dataset annotation. Experienced in collaborating on frontier AI research projects and optimizing NLP systems across text, audio, and visual datasets. Adept at combining psychological insights with technical rigor to ensure nuanced content alignment and user experience optimization.
Alignerr • Remote
Performed structured evaluation and multilingual annotation for AI model improvement. • Performed structured evaluation and multilingual annotation for AI model improvement. • Conducted audio voice tracking and quality scoring using structured rubrics. • Delivered annotated data for supervised learning and multilingual model alignment.
TELUS International AI • Remote
Designed adversarial prompts and stress tests to uncover LLM vulnerabilities. • Designed adversarial prompts and stress tests to uncover LLM vulnerabilities. • Evaluated prompt injection, system override exploits, and compliance breaches. • Logged jailbreaks with ethical framing and persona-based scenario design.
TELUS International AI • Remote
Evaluated model outputs for safety, bias, misinformation, and ethical violations. • Evaluated model outputs for safety, bias, misinformation, and ethical violations. • Scored responses based on severity using platform-specific taxonomies. • Compiled structured feedback reports with risk flags and mitigation strategies.
Outlier • Remote
Rated and annotated translation quality for linguistic datasets. • Rated and annotated translation quality for linguistic datasets. • Categorized errors by severity, fluency, and comprehension standards. • Generated and reviewed queries and answers for annotation-based tasks. • Assessed relevance and effectiveness of search engine results. • Rated digital advertisements for targeting accuracy and user engagement.
Various Clients
Led behavioral research and survey studies across industry verticals. • Led behavioral research and survey studies across industry verticals. • Delivered data-backed insights to optimize brand messaging and user interaction.
Freelance
Delivered professional multilingual voiceovers for regional media and AI audio datasets. • Delivered professional multilingual voiceovers for regional media and AI audio datasets. • Contributed to culturally-sensitive dataset creation for voice recognition models.
Certificate in Teaching English as a Foreign Language