Cyrus W.

Key Skills

Software

Don't disclose

Top Subject Matter

NLP / LLM alignment and model evaluation (RLHF)

Visual annotation and search relevance labeling for training datasets

Legal Services & Contract Review

Top Data Types

Text

Image

Video

Top Task Types

Segmentation

Freelancer Overview

Data Labeling Specialist & Chatbot Evaluator (Remote AI Operations / Distributed Platform Contract). Brings 5+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Don't disclose, Internal, and Proprietary Tooling. Education includes Advanced Secondary School Certificate, N/A and Independent Certification, N/A. AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation, Rating, and Segmentation.

Labeling Experience

Data Labeling Specialist & Chatbot Evaluator (Remote AI Operations / Distributed Platform Contract)

Don't discloseText

Evaluated large-scale LLM and NLP outputs using structured scoring criteria to support model alignment and quality improvements. Performed comparative assessments and generated clear written rationales for RLHF workflows. • Scored model outputs for constraint adherence, factual accuracy, grammar, tone, and logical coherence. • Executed RLHF tasks with concise structured comparative justifications. • Designed edge-case prompts to test safety protocols, semantic reasoning boundaries, and instruction limits. • Audited and cross-verified dataset items to improve alignment pipeline reliability.

2024 - Present

Data Annotation & Quality Assurance Associate (Independent Crowdsourced Data Contracts)

ImageSegmentation

Performed crowdsourced annotation and quality assurance to support training and retrieval-oriented dataset development. Completed both object-level labeling and evaluation of relevance across dataset items. • Labeled images using segmentation and object bounding box style annotations. • Conducted search relevance ranking as part of textual information labeling. • Audited peer-submitted annotation batches to detect errors and noise. • Followed internal guidelines while maintaining high quality and precision across platforms.

2022 - 2024

Education

Work History