Matthew Shepherd

Key Skills

Software

Label Studio

Top Subject Matter

No subject matter listed

Top Data Types

Text

Image

Top Label Types

Classification

Entity Ner Classification

Bounding Box

Freelancer Overview

I am an AI training specialist with over 8 years of experience in data annotation, RLHF evaluation, and high-volume process execution, with a strong background in financial analysis and options trading. My expertise includes text classification, sentiment analysis, content evaluation, image annotation, and structured data labeling, using tools like Label Studio, Argilla, JSON/YAML, and Python. I have built gold-standard test suites for financial LLMs, developed robust QA frameworks, and consistently deliver 99%+ execution accuracy while adhering strictly to guidelines. I thrive in remote, async environments, excel at fact-checking and rubric-based scoring, and am skilled at scenario design, bias identification, and reasoning verification across both financial and general domains. My commitment to quality and calibration ensures reliable, high-integrity training data for advanced AI systems.

Labeling Experience

AI Training & Data Labeling Projects

Label StudioImageBounding Box

Practicing and qualifying in image annotation tasks including bounding box, polygon, and semantic segmentation for model training projects. Completing workflows with rubric-based scoring and aiming for high throughput and accuracy on multiple platforms. Using structured formats (JSON/YAML) in advanced labeling environments for AI training data. • Performed qualification assessments in text classification and content evaluation. • Used Label Studio for annotation practice and task management. • Built accuracy above target thresholds via repeated test cycles. • Training with DataAnnotation Tech and Outlier in concurrent projects.

Practicing and qualifying in image annotation tasks including bounding box, polygon, and semantic segmentation for model training projects. Completing workflows with rubric-based scoring and aiming for high throughput and accuracy on multiple platforms. Using structured formats (JSON/YAML) in advanced labeling environments for AI training data. • Performed qualification assessments in text classification and content evaluation. • Used Label Studio for annotation practice and task management. • Built accuracy above target thresholds via repeated test cycles. • Training with DataAnnotation Tech and Outlier in concurrent projects.

2026

Freelance Research & Data Specialist (Data Annotation)

Label StudioTextEntity Ner Classification

Extracted and labeled entities across institutional programs using named entity recognition and classification. Produced ground-truth datasets with clear flags for ambiguity and exclusion rules, managing guideline-compliant outputs. Documented edge cases and created a calibration library of ambiguous examples. • Performed cross-reference verification for data annotation QA. • Built structured research databases for annotation workflow. • Managed task queues and tagging in Notion for labeling projects. • Maintained 100% guideline alignment in complex research labeling.

Extracted and labeled entities across institutional programs using named entity recognition and classification. Produced ground-truth datasets with clear flags for ambiguity and exclusion rules, managing guideline-compliant outputs. Documented edge cases and created a calibration library of ambiguous examples. • Performed cross-reference verification for data annotation QA. • Built structured research databases for annotation workflow. • Managed task queues and tagging in Notion for labeling projects. • Maintained 100% guideline alignment in complex research labeling.

2020

Independent Financial Analyst & Domain Specialist (Data Labeling)

Label StudioTextClassification

Labeled daily options and derivatives data, categorizing trade setups by strategy type and risk using structured taxonomies. Built and curated gold-standard annotation items, error taxonomy, and rationales for options/derivatives data. Developed decision trees and classification frameworks for trade scenarios. • Fact-checked and verified financial claims for RLHF-style data validation. • Maintained structured annotation QA and correction logs. • Created gold-standard, rubric-based datasets with clear rationale. • Led guideline design for financial data labeling and annotation QA.

Labeled daily options and derivatives data, categorizing trade setups by strategy type and risk using structured taxonomies. Built and curated gold-standard annotation items, error taxonomy, and rationales for options/derivatives data. Developed decision trees and classification frameworks for trade scenarios. • Fact-checked and verified financial claims for RLHF-style data validation. • Maintained structured annotation QA and correction logs. • Created gold-standard, rubric-based datasets with clear rationale. • Led guideline design for financial data labeling and annotation QA.

2018

Education

Work History