AI Training & Data Operations
AI Training & Data Operations Portfolio Kennedy
Maina Kinyua
■ Prompt Evaluation (Before vs After)
Before: 'Write about marketing' After: 'Write a 200-word beginner-friendly explanation of digital
marketing with 3 examples.' Impact: Improved clarity, structure, and relevance of AI outputs.
■ LLM Response Ranking Example
Prompt: Explain climate change to a 10-year-old. Ranked outputs based on simplicity, clarity, and
factual correctness. Top-ranked response used simple language and relatable examples.
■ NLP Entity Recognition Example
Text: 'John works at Safaricom in Nairobi.' PERSON: John | ORGANIZATION: Safaricom |
LOCATION: Nairobi
■ Image Annotation Sample
■ NER Annotation Sample
■ Dataset QA Impact
Maintained 100% annotation accuracy, reduced errors, and improved dataset readiness for model
training.