Isaya John

Key Skills

Software

AWS SageMaker

Anno-Mage

Appen

Argilla

Axiom AI

Clickworker

Doccano

Google Cloud Vertex AI

Trilldata Technologies

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Document

Text

Top Task Types

Bounding Box

Classification

Computer Programming Coding

Question Answering

Text Summarization

Freelancer Overview

I am a Software Engineer and AI Safety Specialist with hands-on experience in evaluating AI outputs, structured annotation, and risk detection. My expertise lies in applying harm taxonomies to identify bias, misinformation, and unsafe reasoning in large language model (LLM) outputs, ensuring that AI systems remain safe, trustworthy, and aligned with human values. As Co-Founder & CTO of Afrigotech, I have led projects in educational technology (EdTech), building SaaS platforms that required rigorous evaluation frameworks to maintain content safety and compliance. This background has strengthened my ability to make careful, consistent decisions in ambiguous situations and document reasoning that improves annotation guidelines. In addition to my safety evaluation experience, I bring a strong foundation in software engineering, data systems, and UI/UX testing, allowing me to spot risks that automated tools might miss. I thrive in fast-moving environments where evaluation methods evolve quickly and am motivated to contribute to advancing AI alignment and safety research.

Labeling Experience

LLM Output Evaluation & AI Safety Annotation

LabelboxTextClassificationText Summarization

Contributed to large-scale annotation projects evaluating the safety of AI-generated text outputs. Tasks included classifying and rating responses against safety guidelines, identifying harmful or biased content, testing prompt–response interactions, and providing structured reasoning for ambiguous cases. Applied harm taxonomies consistently across thousands of samples to ensure data quality. Feedback from evaluations was used to refine safety guidelines and improve model alignment.

Contributed to large-scale annotation projects evaluating the safety of AI-generated text outputs. Tasks included classifying and rating responses against safety guidelines, identifying harmful or biased content, testing prompt–response interactions, and providing structured reasoning for ambiguous cases. Applied harm taxonomies consistently across thousands of samples to ensure data quality. Feedback from evaluations was used to refine safety guidelines and improve model alignment.

2023 - 2024

Education

Work History