DICKSON MWITA

I’m an AI Training Specialist and detail-oriented generalist focused on evaluating and annotating large language model outputs. I use structured rubrics and predefined taxonomies to categorize, label, and improve dataset quality while strengthening reasoning quality, factual accuracy, and clarity.

Since 2024, I’ve worked remotely as a Freelance Data Annotator, applying labeling schemas and annotation guidelines to convert raw inputs into consistent, machine-usable outputs. From 2025 onward, as an AI Chatbot Evaluator with Outlier AI (Scale AI), I evaluate and rank LLM responses, flag hallucinations and guideline violations, and deliver actionable written feedback that supports AI safety and performance improvements.

I’m a reliable, self-directed remote contributor who thrives on guideline adherence, attention to detail, and clear communication. I also craft complex prompts that simulate realistic user needs and participate in onboarding and training assessments to maintain quality and eligibility across concurrent projects.