Paul M.

Key Skills

Software

Mercor

Other

Data Annotation Tech

Clickworker

Top Subject Matter

Multilingual NLP (English/Swahili)

RLHF response evaluation

Conversational AI

Top Data Types

Text

Image

Document

Top Task Types

RLHF

Classification

Segmentation

Freelancer Overview

AI Data Annotator & Evaluator | Mercor AI. Brings 3+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Mercor, Other, and Data Annotation Tech. Education includes Master of Science in Artificial Intelligence, Jomo Kenyatta University of Agriculture and Technology (JKUAT) (2024) and Bachelor of Science, Jomo Kenyatta University of Agriculture and Technology (JKUAT) (2021). AI-training focus includes data types such as Text and Image and labeling workflows including RLHF, Classification, and Segmentation.

Labeling Experience

Contract AI Trainer | DataAnnotation.tech

Data Annotation TechImageSegmentation

Labeled computer vision training data including image bounding boxes, segmentation masks, and classification tags for multi-domain use cases. Provided expert feedback on AI-generated code and prose to improve output quality for next-generation LLMs. Maintained strong inter-annotator agreement on shared benchmarking tasks with attention to domain sensitivity across healthcare, e-commerce, and legal document contexts. • Bounding box, segmentation mask, and classification labeling • Domain-sensitive labeling across healthcare/e-commerce/legal • Benchmarking with high inter-annotator agreement • QA-oriented feedback on generated code and prose

2025 - Present

Freelance Data Annotator | Handshake AI

OtherTextClassification

Performed intent classification, sentiment tagging, and entity recognition for conversational AI training datasets. Annotated code snippets and technical responses to support training and fine-tuning of code generation models. Applied rubrics to assess helpfulness, harmlessness, and honesty across multiple task categories while coordinating to keep labels consistent. • Intent classification and sentiment tagging • Entity recognition for conversational context • Rubric-based evaluation (helpfulness/harmlessness/honesty) • Code snippet and technical response annotation

2025 - Present

AI Data Annotator & Evaluator | Mercor AI

MercorTextRLHF

Evaluated and ranked AI-generated responses for quality, factual accuracy, and coherence to support RLHF pipelines. Annotated multilingual text datasets (English and Swahili) using consistent guidelines to improve NLP performance for low-resource language tasks. Identified edge cases and ambiguous examples and provided structured feedback to refine team labeling rules. • Evaluated response quality and coherence for RLHF • Ranked outputs by factual accuracy • Flagged ambiguities and guideline edge cases • Maintained high acceptance rate and daily throughput targets

2024 - Present

Education

Work History