Al Morgan

Key Skills

Software

AWS SageMaker

LabelImg

Label Studio

SuperAnnotate

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Text

Top Task Types

Action Recognition

Audio Recording

Text Generation

Freelancer Overview

I have several years of experience working with AI training data and linguistic annotation, particularly in speech recognition and natural language processing. I’ve contributed to projects for IBM Watson, Google Assistant, and Nuance Communications, where I labeled and analyzed text and audio data, evaluated AI responses, and helped refine both ASR and text-to-speech models. I’m comfortable following complex guidelines, assessing tone and accuracy, and ensuring consistent, high-quality outputs. My background in computational linguistics and phonetics gives me a strong foundation for understanding how language and sound interact—something that’s especially useful when working with AI models. I’m also a C2-level English writer and editor, which allows me to produce, review, and fact-check content across a wide range of topics while maintaining clarity and precision.

Labeling Experience

Speech Recognition Data Annotation for Multilingual ASR Models

LabelboxTextClassificationText Generation

Contributed to the development of IBM Watson’s Speech-to-Text API by annotating and evaluating multilingual speech datasets. Tasks included phonetic segmentation, utterance classification, and error evaluation to enhance automatic speech recognition (ASR) performance. Collaborated with NLP engineers to refine pronunciation modeling, prosody analysis, and data normalization. The project involved labeling over 100,000 audio segments in English and related dialects. Maintained rigorous QA standards and consistency checks to ensure 98%+ annotation accuracy.

Contributed to the development of IBM Watson’s Speech-to-Text API by annotating and evaluating multilingual speech datasets. Tasks included phonetic segmentation, utterance classification, and error evaluation to enhance automatic speech recognition (ASR) performance. Collaborated with NLP engineers to refine pronunciation modeling, prosody analysis, and data normalization. The project involved labeling over 100,000 audio segments in English and related dialects. Maintained rigorous QA standards and consistency checks to ensure 98%+ annotation accuracy.

2021 - 2025

Education

Work History