Haichuan Z.

Key Skills

Software

AWS SageMaker

Label Studio

Scale AI

Labelbox

CVAT

Roboflow

Top Subject Matter

Machine learning research with speech/ML data pipelines

Conversational Mandarin audio recording for interview content

Top Data Types

Audio

Image

Top Task Types

Transcription

Audio Recording

Freelancer Overview

While I have not worked directly in a formal data labeling role, my background equips me well for AI training data work across multiple domains. My Master of Computing at the Australian National University — covering Machine Learning, Statistical Machine Learning, and Computer Vision — gave me a strong understanding of how annotation quality and consistency directly shape model performance. Beyond technical grounding, my undergraduate training in broadcast media production sharpened my attention to detail, my ability to make consistent judgments on ambiguous material, and my capacity to evaluate content against clear quality standards. In 2017, I also produced a short film selected for the Cannes Film Festival Short Film Corner — a project that demanded rigorous quality control under real constraints. I am a fast learner who thrives in structured, detail-oriented workflows and brings both analytical thinking and creative sensibility to every task.

Labeling Experience

Computing Project — ML Research, Australian National University (2022–2025)

AudioTranscription

The Computing Project — ML Research involved preparing and curating speech-related datasets for machine learning experiments. Work included data preprocessing and quality-focused handling of labeled data that affects downstream model performance. The project emphasized how dataset labeling quality can directly determine results when training and evaluating models. • Performed data curation and preprocessing • Supported labeled dataset preparation for model evaluation • Focused on dataset quality and its impact on model performance • Applied ML techniques within an end-to-end research workflow

2022 - 2025

Education

Work History