Haoyue W.

Key Skills

Software

Labelbox

CVAT

LabelImg

Supervisely

Roboflow

Top Subject Matter

Healthcare diagnostics and clinical data labeling

Business operations and sales analytics datasets

Smart manufacturing documentation and product requirements

Top Data Types

Text

Document

Image

Top Task Types

Diagnosis

Classification

Entity (NER) Classification

Mapping

Freelancer Overview

Clinical Data Annotation Specialist — DeanDiagnostics Technology Group. Brings 4+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal, Proprietary Tooling, and N. Education includes Master of Science, UNSW Sydney (QS#19) (2026) and Bachelor of Science, Zhejiang University of Science and Technology (2024). AI-training focus includes data types such as Medical, DICOM, and Text and labeling workflows including Diagnosis, Classification, and Entity (NER) Classification.

Labeling Experience

LLM Response Quality Annotation & Evaluation — Independent Research

Text

Built a 5-dimension LLM response quality annotation rubric covering factual accuracy, reasoning, instruction-following, coherence, and safety. Annotated 500+ prompt–response pairs to evaluate LLM outputs against the rubric dimensions. Achieved Cohen’s kappa above 0.82 across all dimensions, indicating strong labeling consistency. • Defined rubric criteria for multi-dimensional LLM output evaluation • Performed prompt–response annotation aligned to quality dimensions • Conducted/managed rater agreement scoring using Cohen’s κ • Evaluated LLM responses using structured rubric-based judgments

2025 - Present

Clinical Data Annotation Specialist — DeanDiagnostics Technology Group

Diagnosis

Designed eCRF annotation schema for 8+ disease domains, specifying labeling dimensions, field-level validation rules, and edge-case handling guidelines for consistent team-wide use. Labeled and QA-reviewed 10,000+ structured clinical records to support AI-assisted diagnostic pipeline performance. Managed multi-annotator workflows and maintained inter-annotator agreement above 90% through calibration sessions and ambiguity resolution. • Created CRF schema elements aligned to disease-specific diagnostic labeling needs • Performed systematic batch audits to reduce labeling inconsistency by 15% • Coordinated annotator calibration to align decisions on ambiguous cases • Executed QA review loops to curate high-quality ground truth

2025 - 2026

Digital Product Intern — Data & Documentation (H3C / New H3C Group)

DocumentEntity (NER) Classification

Structured on-site smart manufacturing user research findings into annotated product requirement documents. Organized the resulting requirements into document-ready formats suitable for product documentation and downstream AI/document workflows. Built a Power BI dashboard to track core manufacturing KPIs. • Converted research findings into annotated product requirement documentation • Structured requirements for clarity and consistency across documentation • Supported data-driven product tracking through KPI reporting • Produced annotated documents from real-world facility research

2024 - 2025

Data Annotation & Analysis Intern — MaxMaxis Automation Systems (Jiaxing)

TextClassification

Cleaned and annotated 12 months of sales and operational data using time-series event labels and categorical tags to produce ML-ready training sets. Curated ground-truth label sets for supervised learning pipelines and verified label accuracy via statistical spot-checks and cross-validation review before model handoff. Delivered labeled datasets supporting multiple client projects. • Implemented time-series event labeling and categorical tagging for structured records • Conducted statistical sampling and cross-validation checks to ensure label correctness • Prepared training datasets for three separate client projects • Supported supervised learning pipelines through ground-truth curation

2023 - 2024

Remote Health Equity Data Annotation & Spatial Analysis — UNSW × Western NSW Local Health District

Mapping

Integrated multi-source public health datasets with geo-annotation and accessibility/service-coveragelabels for spatial analytics. Produced QA-validated datasets to support predictive resource-allocation modeling. Delivered interactively consumable spatial dashboard outputs for stakeholder analysis. • Built geospatial datasets with accessibility and service coverage labels • Performed QA validation to ensure dataset readiness for modeling • Supported predictive resource-allocation workflows using labeled data • Enabled interactive spatial dashboard creation from geo-annotated data

2024

Show more

Education

Work History