Taye Omisakin

Key Skills

Software

AWS SageMaker

CVAT

Data Annotation Tech

Dataloop

Encord

Labelbox

OneForma

Redbrick AI

Scale AI

SuperAnnotate

V7 Labs

Other

OpenCV AI Kit (OAK)

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Video

Top Task Types

Audio Recording

Bounding Box

Data Collection

Evaluation/Rating

Tracking

Freelancer Overview

I have over five years of hands-on experience in data labeling and AI training, working across text, image, and video datasets for enterprise-level machine learning projects. My work includes annotating complex visual data, performing OCR transcription, building structured datasets, and evaluating LLM-generated responses for accuracy, reasoning, and instruction-following. I consistently maintain 95–98% quality scores by following detailed guidelines, applying strong analytical skills, and ensuring annotation consistency. I’ve contributed to multiple AI development teams by supporting multimodal dataset preparation, refining annotation guidelines, and performing thorough QA reviews. My experience spans tools like Labelbox, Scale AI, SuperAnnotate, CVAT, and Amazon SageMaker, making me versatile and capable of adapting quickly to new workflows and annotation requirements.

Labeling Experience

OCR Transcription & Document Annotation for NLP Systems

Data Annotation TechDocumentText SummarizationData Collection

Transcribed text from scanned documents and annotated key entities (names, dates, organizations). Worked on data cleaning, text correction, and classification for training an OCR-based NLP system used for automated document processing.

Transcribed text from scanned documents and annotated key entities (names, dates, organizations). Worked on data cleaning, text correction, and classification for training an OCR-based NLP system used for automated document processing.

2022 - 2025

Bounding Box & Object Detection for Autonomous Vehicles

Opencv AI Kit OakVideoBounding BoxTracking

Annotated street scenes for self-driving car datasets, including bounding boxes for pedestrians, vehicles, traffic signs, and lane markings. Performed consistency checks, corrected overlapping box errors, and followed strict pixel-level accuracy guidelines. Contributed to safer real-world navigation models.

Annotated street scenes for self-driving car datasets, including bounding boxes for pedestrians, vehicles, traffic signs, and lane markings. Performed consistency checks, corrected overlapping box errors, and followed strict pixel-level accuracy guidelines. Contributed to safer real-world navigation models.

2021 - 2023

Image Classification for Retail Product Recognition

CVATImageBounding BoxPolygon

Labeled over 30,000 retail product images for an AI model used by an e-commerce platform. Tasks included categorizing products, tagging attributes (brand, color, size), and identifying mislabeled items. Achieved 98% accuracy and helped improve object detection performance in live customer-facing applications.

Labeled over 30,000 retail product images for an AI model used by an e-commerce platform. Tasks included categorizing products, tagging attributes (brand, color, size), and identifying mislabeled items. Achieved 98% accuracy and helped improve object detection performance in live customer-facing applications.

2021 - 2023

Education

Work History