Michael Peterson

Key Skills

Software

Anno-Mage

Appen

CVAT

Data Annotation Tech

iMerit

Labelbox

Lionbridge

Prodigy

Sloth

SuperAnnotate

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Audio Recording

Bounding Box

Classification

Entity Ner Classification

Segmentation

Freelancer Overview

I am an experienced AI training data specialist with over 5 years of hands-on work in data labeling, transcription, and large language model (LLM) evaluation. I’ve contributed to projects across natural language processing (NLP), computer vision, and audio processing—helping train AI systems through precise text, image, video, and audio annotations. I have worked with industry-standard tools like Labelbox, CVAT, SuperAnnotate, and VIA, and am skilled in transcription platforms such as Verbit, Rev, and TELUS. My strengths lie in producing high-quality data aligned with strict guidelines, working efficiently under deadlines, and quickly adapting to new task types and ontologies. I’ve handled tasks such as LLM prompt evaluation, entity recognition, content classification, bounding box labeling, timestamped transcription, and audio diarization. With a deep understanding of both the technical and linguistic nuances of data preparation, I’m well-equipped to support scalable, high-impact AI development projects.

Labeling Experience

Image Annotation for Object Detection & Classification

CVATImageBounding BoxPolygon

Labeled thousands of images for object detection and classification tasks across multiple industries. Annotated product images for e-commerce catalogs, and street scenes for autonomous vehicle datasets using bounding boxes and class tags. Ensured label accuracy and consistency according to detailed project guidelines. Also performed QA review on peer annotations to flag edge cases, misclassifications, and labeling errors. Maintained accuracy above 97% over the course of the project.

Labeled thousands of images for object detection and classification tasks across multiple industries. Annotated product images for e-commerce catalogs, and street scenes for autonomous vehicle datasets using bounding boxes and class tags. Ensured label accuracy and consistency according to detailed project guidelines. Also performed QA review on peer annotations to flag edge cases, misclassifications, and labeling errors. Maintained accuracy above 97% over the course of the project.

2023 - 2024

Multi-Modal Annotation for AI Model Training

LabelboxTextEntity Ner ClassificationClassification

Worked on high-volume text annotation and evaluation tasks to support the training and fine-tuning of a large language model (LLM). Tasks included prompt creation, entity tagging, sentiment labeling, and ranking AI-generated responses for fluency, relevance, and factual accuracy. Followed evolving annotation guidelines and maintained a 98%+ quality score across thousands of submissions. Provided written feedback to improve model behavior and reduce bias in generative outputs.

Worked on high-volume text annotation and evaluation tasks to support the training and fine-tuning of a large language model (LLM). Tasks included prompt creation, entity tagging, sentiment labeling, and ranking AI-generated responses for fluency, relevance, and factual accuracy. Followed evolving annotation guidelines and maintained a 98%+ quality score across thousands of submissions. Provided written feedback to improve model behavior and reduce bias in generative outputs.

2023 - 2024

Audio Transcription & Diarization for Speech-to-Text AI Training

Internal Proprietary ToolingAudioSegmentationQuestion Answering

Contributed to a multilingual speech-to-text AI model training project by delivering high-accuracy transcriptions from diverse audio sources, including interviews, podcasts, and webinars. Tasks involved verbatim and clean transcription, speaker labeling, insertion of timecodes, and proofreading. Managed large audio files with multiple speakers and varying accents. Maintained quality benchmarks above 96% and consistently met daily throughput targets. Used Audacity for segmentation and quality control.

Contributed to a multilingual speech-to-text AI model training project by delivering high-accuracy transcriptions from diverse audio sources, including interviews, podcasts, and webinars. Tasks involved verbatim and clean transcription, speaker labeling, insertion of timecodes, and proofreading. Managed large audio files with multiple speakers and varying accents. Maintained quality benchmarks above 96% and consistently met daily throughput targets. Used Audacity for segmentation and quality control.

2022 - 2024

Education

Work History