Elvis Owido

Key Skills

Software

CloudFactory

Data Annotation Tech

Datature

Google Cloud Vertex AI

Labelbox

Remotasks

Sama

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Text

Top Label Types

Bounding Box

Computer Programming Coding

Cuboid

Evaluation Rating

Segmentation

Freelancer Overview

Elvis Owido is an accomplished AI professional with extensive experience in data labeling, annotation, and AI training data systems. With a proven record across organizations such as Invisible Technologies, Outlier, and DataAnnotation Tech, he has led end-to-end data initiatives that fuel high-performing machine learning and natural language processing models. His expertise covers data collection, preprocessing, annotation, and validation ensuring accuracy, consistency, and compliance with ethical AI standards. Notably, he improved image classification accuracy by 20% through a deep learning algorithm and enhanced NLP chatbot responsiveness by transitioning to a GPT-based framework. Elvis combines strong technical proficiency with strategic insight. He is highly skilled in Python (TensorFlow, Keras), Java, SQL, and modern data frameworks such as Hadoop, Spark, and Kafka. Experienced in cloud environments including AWS and Google Cloud, he leverages scalable infrastructures to optimize model training workflows. Holding a Master’s degree in Artificial Intelligence with a focus on reinforcement learning, he brings a deep understanding of how well-structured data underpins AI success. His leadership, precision, and technical versatility position him as a valuable contributor to projects involving AI data labeling, model optimization, and intelligent automation.

Labeling Experience

Image Data Labeling for Computer Vision Models

Data Annotation TechImageBounding BoxSegmentation

Contributed to large-scale image annotation projects at Data Annotation Tech, focusing on improving object detection and image classification accuracy for AI-powered systems. I applied bounding boxes and segmentation masks to label thousands of industrial and vehicular images used for model training. The resulting dataset powered a deep learning model that achieved a 20% improvement in classification accuracy and reduced misidentification errors. Quality was maintained through multi-stage verification, achieving 98%+ labeling precision. I collaborated with AI engineers to align annotation categories with model training objectives and ensure consistency across datasets.

Contributed to large-scale image annotation projects at Data Annotation Tech, focusing on improving object detection and image classification accuracy for AI-powered systems. I applied bounding boxes and segmentation masks to label thousands of industrial and vehicular images used for model training. The resulting dataset powered a deep learning model that achieved a 20% improvement in classification accuracy and reduced misidentification errors. Quality was maintained through multi-stage verification, achieving 98%+ labeling precision. I collaborated with AI engineers to align annotation categories with model training objectives and ensure consistency across datasets.

2024

Image and Text Data Labeling for AI Model Training

SamaImageBounding BoxEntity Ner Classification

Led and contributed to multiple large-scale data labeling and annotation projects at Sama. Tasks involved creating high-quality labeled datasets for computer vision and NLP models. For image data, I applied bounding boxes and segmentation techniques to identify objects and scenes, achieving a 20% improvement in image classification accuracy for a production-level deep learning model. For text data, I performed entity recognition, intent classification, and chatbot response evaluation to enhance GPT-based conversational AI systems resulting in a 50% reduction in escalation rates. Each project adhered to rigorous quality assurance standards, including double-blind reviews and accuracy benchmarks above 98%. Collaborated with cross-functional teams to ensure annotation consistency and ethical data handling in alignment with international standards.

Led and contributed to multiple large-scale data labeling and annotation projects at Sama. Tasks involved creating high-quality labeled datasets for computer vision and NLP models. For image data, I applied bounding boxes and segmentation techniques to identify objects and scenes, achieving a 20% improvement in image classification accuracy for a production-level deep learning model. For text data, I performed entity recognition, intent classification, and chatbot response evaluation to enhance GPT-based conversational AI systems resulting in a 50% reduction in escalation rates. Each project adhered to rigorous quality assurance standards, including double-blind reviews and accuracy benchmarks above 98%. Collaborated with cross-functional teams to ensure annotation consistency and ethical data handling in alignment with international standards.

2019 - 2022

Education

Work History