Joseph Mwaniki

Key Skills

Software

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Label Types

Evaluation Rating

Freelancer Overview

I am a detail-oriented professional with a background in forensic science and hands-on experience in data annotation, labeling, and QA review for AI training data. My work as a freelance AI and QA contributor has involved evaluating AI-generated text, annotating datasets, fact-checking outputs, and providing structured feedback to refine machine learning models, primarily in natural language processing tasks. I am skilled in following complex guidelines, ensuring data accuracy, and using tools such as Python (NumPy, Pandas), Excel, and various online platforms. My analytical mindset, experience in digital forensics, and ability to work independently in remote settings enable me to deliver high-quality, reliable data for AI development and research projects. In terms of hands-on experience, I have worked on large language model (LLM) evaluation projects where I compared and ranked AI-generated responses based on reasoning accuracy, factual correctness, safety, and clarity. I have also participated in text classification and sentiment analysis projects, annotating datasets for intent detection and content moderation while maintaining strict adherence to annotation guidelines. Additionally, I have conducted fact-checking and quality assurance reviews to identify hallucinations, inconsistencies, and policy violations in AI outputs. My responsibilities have included prompt refinement and edge-case testing to assess instruction-following capabilities, as well as structured data cleaning and validation using Python and Excel to ensure dataset integrity before final submission.

Labeling Experience

AI Trainer / Data Annotator – Outlier AI & Remotask

Scale AITextEvaluation Rating

As an AI Trainer and Data Annotator at Outlier AI and Remotask, I evaluated and ranked AI-generated responses for accuracy, clarity, safety, and reasoning quality. I created and refined prompts spanning several domains to improve AI performance. Structured feedback was provided to optimize AI intelligence and analytical model comparisons were performed. • Evaluated AI-generated text responses across multiple quality metrics • Created and refined prompts for diverse subject areas • Provided high-quality structured feedback to enhance AI models • Collaborated in remote teams while maintaining deadlines and standards

As an AI Trainer and Data Annotator at Outlier AI and Remotask, I evaluated and ranked AI-generated responses for accuracy, clarity, safety, and reasoning quality. I created and refined prompts spanning several domains to improve AI performance. Structured feedback was provided to optimize AI intelligence and analytical model comparisons were performed. • Evaluated AI-generated text responses across multiple quality metrics • Created and refined prompts for diverse subject areas • Provided high-quality structured feedback to enhance AI models • Collaborated in remote teams while maintaining deadlines and standards

2022

Education

Work History