Vinton Juniour

Key Skills

Software

AWS SageMaker

Appen

Clickworker

CrowdSource

Data Annotation Tech

OneForma

Remotasks

Toloka

Telus

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Document

Text

Top Task Types

Bounding Box

Classification

Computer Programming Coding

Data Collection

Text Summarization

Freelancer Overview

I’m a skilled and reliable Multimodal Data Annotation Specialist with hands-on experience labeling text, audio, image, and video datasets to train cutting-edge AI systems. Over the past two years, I’ve completed high-accuracy labeling projects on platforms like Fiverr, Upwork, Remotasks, and UHRS, supporting domains such as computer vision, natural language processing, and speech recognition. I’m proficient in a range of annotation tools including Labelbox, CVAT, Roboflow, and Label Studio, and I’ve worked on projects involving bounding box annotation, NER tagging, text classification, audio transcription, and LLM evaluation. My strengths include fast adaptation to new guidelines, consistent quality, and a deep commitment to improving AI through precise, clean, and scalable datasets.

Labeling Experience

UHRS – Text Categorization & Search Evaluation

ClickworkerTextEntity Ner ClassificationClassification

Performed query intent classification, relevance scoring, and entity recognition across user-generated queries. Ensured consistency across batches and provided natural language feedback to improve LLM training and result quality.

Performed query intent classification, relevance scoring, and entity recognition across user-generated queries. Ensured consistency across batches and provided natural language feedback to improve LLM training and result quality.

2022

Remotasks – Computer Vision Annotation

Scale AIVideoBounding BoxPolygon

Labeled street-level images and dashcam footage for self-driving car datasets. Tasks included identifying and drawing bounding boxes and polygons around vehicles, pedestrians, traffic signs, and lane lines. Maintained over 97% accuracy across thousands of tasks through regular QA reviews and feedback loops.

Labeled street-level images and dashcam footage for self-driving car datasets. Tasks included identifying and drawing bounding boxes and polygons around vehicles, pedestrians, traffic signs, and lane lines. Maintained over 97% accuracy across thousands of tasks through regular QA reviews and feedback loops.

2023 - 2023

Education

Work History