Dorcas M.

Key Skills

Software

Don't disclose

Labelbox

Top Subject Matter

Healthcare Data & Medical Documentation

Language Domain Expertise

digital communication

Top Data Types

Text

Top Task Types

RLHF

Red Teaming

Freelancer Overview

AI Trainer & AI Quality Specialist (Freelance) - AI content evaluation and RLHF quality work. Brings 9+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose and Labelbox. Education includes Bachelor of Science, University of Nairobi and Good Clinical Practice Certification, NIDA Clinical Trials Network. AI-training focus includes data types such as Text and labeling workflows including RLHF, Entity (NER), and Red Teaming.

Labeling Experience

AI Evaluation and Quality Specialist (Freelance) - Freelance

LabelboxTextRLHF

You evaluate AI-generated responses for accuracy, completeness, safety, tone, and contextual relevance to support higher-quality model outputs. You perform RLHF assessments and provide structured feedback to improve model alignment, reduce hallucinations, and mitigate unsafe or biased outputs. You also support multilingual projects and collaborate on improving evaluation guidelines and annotation workflows across cross-functional teams. • Conduct factuality, safety, and contextual relevance reviews • Perform RLHF evaluation and preference-quality checks • Identify bias, misinformation, and culturally sensitive errors • Audit quality and provide structured improvement feedback

2021 - Present

AI Trainer & AI Quality Specialist (Freelance) - Safety/red-team evaluation of LLM responses

Don't discloseTextRed Teaming

You executed trust & safety evaluation activities by performing red-teaming style review of model responses for unsafe behavior and policy compliance. You identified bias, toxicity/unsafe language, misinformation, ambiguity, and culturally insensitive content and escalated structured findings. You applied evaluation frameworks and human feedback to improve moderation and safety behavior in LLM outputs. • Detected harmful/unsafe content and policy violations in LLM responses. • Identified bias, ambiguity, misinformation, and contextual safety risks. • Performed hallucination detection and quality checks for factual reliability. • Contributed to AI safety improvements through structured evaluation feedback.

2021 - Present

AI Trainer & AI Quality Specialist (Freelance) - Text annotation for NLP (sentiment/intent/NER/conversational AI)

LabelboxText

You annotated large text datasets for NLP projects including conversational AI and linguistic analysis. You labeled information such as sentiment, intent, entities, and conversational attributes to support downstream model training and evaluation. You ensured label quality by applying structured criteria and participating in quality assurance improvements. • Labeled sentiment, intent classification, entity recognition, and conversational AI data. • Supported multilingual dataset creation for English and Swahili. • Performed linguistic evaluation and behavioral analysis labeling. • Audited annotation quality and helped maintain QA standards.

2021 - Present

AI Trainer & AI Quality Specialist (Freelance) - AI content evaluation and RLHF quality work

Don't discloseTextRLHF

You performed RLHF-style assessments to improve model alignment and response quality by evaluating human preferences and response behavior. You audited AI-generated outputs for factual accuracy, safety, completeness, tone, and contextual relevance across healthcare and multilingual communication settings. You provided structured feedback and identified failure modes such as hallucinations, harmful content, and policy violations to support model improvement. • Evaluated AI responses for factual accuracy, completeness, safety, tone, and context. • Conducted RLHF assessments and human feedback/rating to improve alignment. • Detected hallucinations, harmful content, bias, and contextual errors. • Supported guideline and workflow improvements with cross-functional teams.

2021 - Present

Education

Work History