Kun W.

Key Skills

Software

Don't disclose

Top Subject Matter

AI response/code evaluation for Python tasks

Text annotation quality assurance and label consistency

Top Data Types

Text

Top Task Types

Classification

Freelancer Overview

Python Code Evaluation Toolkit (AI training readiness). Core strengths include Don't disclose. Education includes Bachelor of Engineering, Zhejiang University (2025) and Certificate, N/A (2026). AI-training focus includes data types such as Computer Code, Programming, and Text and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

Text Annotation Quality Checker (AI training readiness)

Don't discloseTextClassification

Developed a text annotation quality-control utility to validate annotation fields and enforce label consistency. Detected missing annotation fields, unsupported labels, and duplicate record IDs to improve dataset reliability. Measured inter-annotator agreement for shared task IDs to monitor consistency. • Validated required fields and schema compliance • Flagged unsupported labels and duplicate record identifiers • Computed exact inter-annotator agreement for shared items • Confirmed validation and agreement logic with unit tests

2026 - Present

Python Code Evaluation Toolkit (AI training readiness)

Don't disclose

Built a Python-based evaluator to run input/expected-output test cases and record outcomes for reproducible code review. Assessed code behavior against explicit requirements and surfaced incorrect logic, wrong answers, and exception paths. Produced structured results summaries to support evidence-based feedback. • Executed unit-test scenarios for correct outputs, wrong answers, and exceptions • Logged pass/fail/exception outcomes for traceable evaluation • Generated aggregated pass-rate and incorrect-output details • Verified evaluator behavior using multiple automated test cases

2026 - Present

Education

Work History