Michael A.

Key Skills

Software

AWS SageMaker

Clickworker

Appen

Anno-Mage

Argilla

Axiom AI

Top Subject Matter

Healthcare-Medical Records and Patient Data

Finance-Financial Analysis and Fraud Detection

E-Commerce-Product Categorization and Customer Support

Top Data Types

Video

Image

Document

Top Task Types

Object Detection

Cuboid

Bounding Box

Segmentation

Classification

Entity (NER) Classification

Point/Key Point

Polyline

Text Generation

Question Answering

Text Summarization

RLHF

Fine-tuning

Red Teaming

Polygon

Evaluation/Rating

Transcription

Computer Programming/Coding

Data Collection

Function Calling

Prompt + Response Writing (SFT)

Freelancer Overview

I am a Stanford University Graduate In Computer Science and Engineering(Software, Computer and Data) in which over the past 6 years, I have specialized in curating high-quality training datasets that directly improve AI model accuracy and performance. My experience spans a diverse range of data modalities, including natural language processing (NLP), text classification, and human preference ranking (RLHF) for large language models of which rather than just hitting volume targets, I focus heavily on precision and edge-case documentation, consistently maintaining a near-perfect accuracy rating across complex taxonomy guidelines. I am also highly adaptable, comfortable working with industry-standard annotation platforms, and thrive in fast-paced environments where data requirements evolve alongside the model's needs. What sets me apart is my proactive approach to data quality. I don't just label data; I actively identify ambiguities in instruction sets that lead to inter-annotator disagreement. By flagging these inconsistencies early and collaborating on guideline calibration, I have previously helped teams significantly reduce label conflicts and accelerate dataset delivery. I bring a meticulous eye for detail, a strong grasp of how high-quality data impacts machine learning outputs, and the reliable communication skills necessary to deliver production-ready training data.

Education

Work History