AI Training & QA Projects
All work shown here is my original creation. While I cannot share client or proprietary data due to NDAs, these public samples reflect the same quality, attention to detail, and professional approach I apply to every project.
PROJECT 1:
📊 COMPARATIVE EVALUATION OF AI RESPONSES
Evaluated AI-generated responses from ChatGPT, Claude, and Gemini across multiple prompt categories using a consistent 5-criteria rubric (Accuracy, Clarity, Completeness, Tone, Value). Documented strengths, weaknesses, and inconsistencies between models, providing actionable feedback for improvement. Demonstrates systematic QA methodology, attention to detail, and comparative analysis skills essential for AI training and supervised fine-tuning work.
PROJECT 2: 🐛 BUG HUNTING & ISSUE DOCUMENTATION
Conducted exploratory testing on public web applications to identify, document, and reproduce software bugs. Created detailed bug reports following industry standards—including steps to reproduce, expected vs. actual behavior, environment details, and severity assessment. Uncovered issues ranging from checkout errors and mobile navigation failures to form validation gaps and search function inconsistencies. Demonstrates sharp attention to detail, a systematic testing approach, and professional bug documentation skills currently applied on uTest and other QA platforms.
PROJECT 3: ⚡ ENGLISH–SWAHILI LINGUISTIC REVIEW
Evaluated AI translation tools (Google Translate, ChatGPT, and Claude) against 20 common phrases across customer service, marketing, technical, and conversational contexts. Identified issues ranging from literal translations that sound unnatural to completely wrong meanings. Provided improved versions that reflect authentic Kenyan Swahili—considering cultural context, tone, and natural flow. Demonstrates native-level linguistic expertise, attention to nuance, and the critical value of human review in AI training and localization projects.
PROJECT 4: PROMPT ENGINEERING PORTFOLIO
Demonstrates the art of crafting effective prompts that generate high-quality AI responses. Features before/after examples showing how strategic refinements—adding audience context, tone specifications, structural constraints, and clear objectives—transform generic outputs into targeted, valuable content. Includes samples across content creation, creative writing, problem-solving, and translation tasks. Showcases the critical skill of iterative prompt design used daily in AI training, supervised fine-tuning, and quality assurance work.