Busonma Edo

Key Skills

Software

Mindrift

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Label Types

Question Answering

Text Generation

Emotion Recognition

Translation Localization

Freelancer Overview

I have hands-on experience in AI training and data annotation through my work with Outlier and Appen, where I contributed to training large language models by evaluating outputs, refining prompts, labeling datasets, and ensuring responses met quality and policy standards. My role required careful attention to detail, consistency in applying guidelines, and the ability to assess tone, factual accuracy, and contextual relevance. I regularly worked with structured rubrics to provide clear, evidence-based feedback that improved model performance. What sets me apart is my analytical background in law and leadership within a healthcare environment, which strengthens my precision, critical thinking, and quality control skills. I am comfortable interpreting complex instructions, identifying subtle errors or biases, and delivering well-reasoned evaluations under tight deadlines. My approach combines technical accuracy with a human-centred perspective, ensuring AI systems are both reliable and aligned with real-world user needs.

Labeling Experience

Large Language Model Response Evaluation & Annotation

MindriftTextQuestion AnsweringText Generation

Contributed to training and improving large language models by evaluating AI-generated responses for accuracy, coherence, tone, safety, and policy compliance. Applied structured rubrics to rank outputs, identify hallucinations, flag policy violations, and provide detailed written justifications to support model refinement. Completed high-volume annotation tasks while maintaining strict quality benchmarks and turnaround times.

Contributed to training and improving large language models by evaluating AI-generated responses for accuracy, coherence, tone, safety, and policy compliance. Applied structured rubrics to rank outputs, identify hallucinations, flag policy violations, and provide detailed written justifications to support model refinement. Completed high-volume annotation tasks while maintaining strict quality benchmarks and turnaround times.

2025 - 2025

Education

Work History