Amos Muriithi

In my remote contract work, I evaluate AI-generated responses for factual accuracy, relevance, completeness, and instruction following. I compare alternative responses and write structured rationales that identify strengths, weaknesses, and factual discrepancies.

I also review outputs for unsupported claims and logical inconsistencies, and assess English–Kiswahili translations for meaning, grammar, and cultural appropriateness. I analyze multi-turn conversations for context retention and develop prompts to test reasoning and formatting compliance.

As a Linguistic Content Specialist and Educational Assessment Support, I develop assessment criteria and evaluate written work against established rubrics. My work also includes analyzing English and Kiswahili language structures and preparing educational materials and reference resources.

On my Multilingual AI Evaluation and Stress Testing project, I designed a collection of 60+ prompts covering instruction following, complex constraints, logical negation, and English–Kiswahili translation. I also developed a five-point response-comparison approach and criteria for relevance, accuracy, instruction adherence, clarity, and completeness.