Presh Gallant

Key Skills

Software

Label Studio

CVAT

Argilla

Datasaur

Top Subject Matter

E-commerce Product Data & Taxonomy

Retail Domain Expertise

Social Media

Top Data Types

Image

Text

Document

Top Task Types

Bounding Box

Classification

Freelancer Overview

AI Data Annotation Specialist (Contract). Brings 11+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Label Studio, CVAT, and Argilla. Education includes Bachelor of Commerce, University of Toronto (2016). AI-training focus includes data types such as Image and Text and labeling workflows including Bounding Box and Classification.

Labeling Experience

AI Data Annotation Specialist (Contract)

Label StudioImageBounding Box

Labeled and segmented over 60,000 product images and descriptions for e-commerce AI models. Created bounding boxes and semantic segmentation masks, improving model visual recognition and search results. Carried out LLM fine-tuning with preference labeling and RLHF-style evaluation of 8,000+ prompt-response pairs. • Achieved a labeling accuracy rate of 98.7% • Reduced QA rework cycles by 35% • Labeled both image and text data for multimodal tasks • Used Label Studio and CVAT extensively for annotation

Labeled and segmented over 60,000 product images and descriptions for e-commerce AI models. Created bounding boxes and semantic segmentation masks, improving model visual recognition and search results. Carried out LLM fine-tuning with preference labeling and RLHF-style evaluation of 8,000+ prompt-response pairs. • Achieved a labeling accuracy rate of 98.7% • Reduced QA rework cycles by 35% • Labeled both image and text data for multimodal tasks • Used Label Studio and CVAT extensively for annotation

2023 - Present

Content Moderation & Data Labeler

ArgillaTextClassification

Annotated more than 45,000 social media posts and comments for sentiment and toxicity detection. Contributed to dataset curation that supports sentiment analysis and brand-safety models in social media contexts. Performed multilingual labeling using Argilla and internal tools to enhance model precision. • Raised model precision by 22% through rubric-based annotation • Managed English and French language labeling tasks • Built labeled benchmark sets for red-teaming and audit purposes • Maintained consistency with rubric and project standards

Annotated more than 45,000 social media posts and comments for sentiment and toxicity detection. Contributed to dataset curation that supports sentiment analysis and brand-safety models in social media contexts. Performed multilingual labeling using Argilla and internal tools to enhance model precision. • Raised model precision by 22% through rubric-based annotation • Managed English and French language labeling tasks • Built labeled benchmark sets for red-teaming and audit purposes • Maintained consistency with rubric and project standards

2021 - 2023

Data Quality Analyst

DatasaurTextClassification

Labeled and classified over 20,000 customer interaction transcripts and transaction records for NLP fraud detection models. Developed annotation guidelines to maintain consistent and reliable labeling. Conducted quality assurance and peer review for chatbot training datasets to ensure high inter-annotator agreement. • Achieved 99% inter-annotator agreement for labeled data • Enhanced fraud-detection and chatbot reliability • Focused on finance and bank customer service transcripts • Ensured data quality control through standardized QA processes

Labeled and classified over 20,000 customer interaction transcripts and transaction records for NLP fraud detection models. Developed annotation guidelines to maintain consistent and reliable labeling. Conducted quality assurance and peer review for chatbot training datasets to ensure high inter-annotator agreement. • Achieved 99% inter-annotator agreement for labeled data • Enhanced fraud-detection and chatbot reliability • Focused on finance and bank customer service transcripts • Ensured data quality control through standardized QA processes

2019 - 2021

Education

Work History