Steven Wang

Key Skills

Software

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Image

Top Label Types

Bounding Box

Polygon

Classification

Freelancer Overview

I bring over a decade of experience designing and deploying advanced AI and machine learning systems, with a strong focus on the full data lifecycle—from data preprocessing and feature engineering to model training, evaluation, and deployment. My expertise includes building scalable data pipelines, developing robust evaluation harnesses for model quality (using metrics like ROUGE and BLEU), and integrating human-in-the-loop feedback for continuous improvement of AI agents. I am highly skilled in Python, TensorFlow, PyTorch, and major cloud ML platforms such as AWS SageMaker and Google Vertex AI. My work has spanned a range of domains, including enterprise applications and internal knowledge retrieval, where I’ve applied advanced embedding models and semantic search to optimize data annotation and labeling workflows. I am passionate about ensuring high-quality, reliable training data and collaborating with cross-functional teams to deliver AI solutions that meet real-world business needs.

Labeling Experience

Image Annotation for Object Detection (Bounding Boxes & Segmentation)

LabelboxImageBounding BoxPolygon

Worked on a large-scale image annotation project focused on object detection and visual classification. Tasks included drawing tight bounding boxes, creating polygon segmentation masks, and assigning multi-class labels based on detailed annotation guidelines. Annotated 20,000+ images across multiple categories while maintaining high accuracy and consistency. Followed strict quality assurance workflows, including peer review, validation checks, and feedback-based corrections to meet precision and recall targets required for machine learning training datasets.

Worked on a large-scale image annotation project focused on object detection and visual classification. Tasks included drawing tight bounding boxes, creating polygon segmentation masks, and assigning multi-class labels based on detailed annotation guidelines. Annotated 20,000+ images across multiple categories while maintaining high accuracy and consistency. Followed strict quality assurance workflows, including peer review, validation checks, and feedback-based corrections to meet precision and recall targets required for machine learning training datasets.

2020 - 2025

Education

Work History