Divine-treasure Ajayi

Key Skills

Software

CVAT

Data Annotation Tech

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Text

Top Task Types

Bounding Box

Classification

Entity Ner Classification

Object Detection

Polygon

Freelancer Overview

Intermediate AI data annotator specializing in LLM evaluation and text analysis for NLP projects. Experienced in labeling text, image, and video data, including bounding boxes, polygons, and frame-level video annotation using CVAT. Skilled in Python (Pandas, NumPy, Matplotlib) for structured data cleaning, preprocessing, and analysis to support AI training workflows. I focus on accuracy, consistency, and high-quality annotations that improve model performance. Passionate about AI training, data-driven insights, and delivering reliable datasets for effective machine learning applications

Labeling Experience

LLM Text Evaluation & Annotation (English)

CVATTextEntity Ner ClassificationClassification

Evaluated AI-generated English text outputs for relevance, coherence, and instruction-following. Classified text samples and annotated key entities using a structured workflow in CVAT to maintain consistency and quality. Ensured all annotations adhered to guidelines for LLM training and evaluation.

Evaluated AI-generated English text outputs for relevance, coherence, and instruction-following. Classified text samples and annotated key entities using a structured workflow in CVAT to maintain consistency and quality. Ensured all annotations adhered to guidelines for LLM training and evaluation.

2025 - 2025

Python Code Annotation & AI-Generated Script Review

LabelboxComputer Code ProgrammingEntity Ner ClassificationClassification

In this project, I annotated and labeled Python scripts to prepare high-quality training datasets for AI models. My tasks included marking function names, variable usage, input/output types, and expected results, as well as validating AI-generated code outputs for correctness and consistency. I utilized Prodigy and Labelbox, along with custom annotation scripts, to ensure structured labeling that aligns with AI training requirements. This workflow contributed to enhanced model performance, accurate function execution prediction, and robust code understanding for AI systems.

In this project, I annotated and labeled Python scripts to prepare high-quality training datasets for AI models. My tasks included marking function names, variable usage, input/output types, and expected results, as well as validating AI-generated code outputs for correctness and consistency. I utilized Prodigy and Labelbox, along with custom annotation scripts, to ensure structured labeling that aligns with AI training requirements. This workflow contributed to enhanced model performance, accurate function execution prediction, and robust code understanding for AI systems.

2024 - 2024

Frame Annotation & Object Tracking

LabelboxVideoObject DetectionTracking

Performed detailed frame-by-frame annotation of video datasets for object tracking and detection tasks. Tracked multiple objects across video sequences, ensuring temporal consistency and precise labeling for each frame. Implemented quality assurance protocols, including verification of frame continuity and consistency checks for object boundaries. Managed annotation of 50 short videos (5–10 minutes each) containing multiple objects, providing datasets structured and ready for AI training. Leveraged Labelbox’s video annotation tools for efficient tracking, labeling, and dataset management.

Performed detailed frame-by-frame annotation of video datasets for object tracking and detection tasks. Tracked multiple objects across video sequences, ensuring temporal consistency and precise labeling for each frame. Implemented quality assurance protocols, including verification of frame continuity and consistency checks for object boundaries. Managed annotation of 50 short videos (5–10 minutes each) containing multiple objects, providing datasets structured and ready for AI training. Leveraged Labelbox’s video annotation tools for efficient tracking, labeling, and dataset management.

2024 - 2024

Image Annotation & Bounding Box Labeling

CVATImageBounding BoxPolygon

Annotated and labeled a large-scale image dataset with 1000+ images for object detection tasks using bounding boxes and polygon annotations. Carefully identified and tagged multiple object categories per image, ensuring consistency and precision across the dataset. Implemented quality control measures, including cross-checking annotations and verifying accuracy for each batch. Optimized workflow by combining CVAT for core annotation and Labelbox for verification and review. Delivered structured and high-quality datasets ready for training and evaluation of computer vision models.

Annotated and labeled a large-scale image dataset with 1000+ images for object detection tasks using bounding boxes and polygon annotations. Carefully identified and tagged multiple object categories per image, ensuring consistency and precision across the dataset. Implemented quality control measures, including cross-checking annotations and verifying accuracy for each batch. Optimized workflow by combining CVAT for core annotation and Labelbox for verification and review. Delivered structured and high-quality datasets ready for training and evaluation of computer vision models.

2024 - 2024

Education

Work History