Ɥǝʇɐqo Ochieng

Key Skills

Software

Clickworker

CloudFactory

CVAT

Data Annotation Tech

Hasty

iMerit

Labelbox

OneForma

Remotasks

Scale AI

V7 Labs

Top Subject Matter

No subject matter listed

Top Data Types

3D Sensor

Image

Text

Top Task Types

Cuboid

Data Collection

Object Detection

Point/Key Point

Polygon

Freelancer Overview

I specialize in high-quality data labeling and AI training workflows across computer vision, natural language processing, and multilingual LLM evaluation. My experience spans annotating medical images for diagnostic AI, labeling retail product datasets for e-commerce platforms, and evaluating large language models in both English and Swahili. I’m proficient in tools like CVAT, Labelbox, and Prodigy, and I understand the importance of consistency, edge-case handling, and annotation guidelines tailored to model objectives. What sets me apart is my ability to combine domain knowledge with precision annotation—whether it's segmenting anatomical structures in radiology scans or curating prompt-response pairs for LLM fine-tuning. I’ve contributed to projects involving autonomous systems, healthcare AI, and multilingual chatbot development, always ensuring that the labeled data meets the highest standards for downstream model performance.

Labeling Experience

3D Sensor Annotation for Autonomous Vehicle Perceptio

CVAT3D SensorBounding BoxPolygon

This project involved annotating LiDAR and camera sensor data to train perception models for autonomous driving systems. I labeled thousands of frames with 3D bounding boxes and cuboids to identify vehicles, pedestrians, cyclists, and static infrastructure. Using CVAT, I applied polygon and segmentation tools to delineate road boundaries, lane markings, and traffic signs with pixel-level precision. The dataset spanned urban, suburban, and highway environments, requiring careful attention to occlusions, motion blur, and edge cases. Quality assurance was a top priority—I followed strict labeling guidelines, participated in regular calibration reviews, and maintained over 98% accuracy across validation checks. My contributions helped improve object tracking and scene understanding for real-time decision-making in autonomous navigation.

This project involved annotating LiDAR and camera sensor data to train perception models for autonomous driving systems. I labeled thousands of frames with 3D bounding boxes and cuboids to identify vehicles, pedestrians, cyclists, and static infrastructure. Using CVAT, I applied polygon and segmentation tools to delineate road boundaries, lane markings, and traffic signs with pixel-level precision. The dataset spanned urban, suburban, and highway environments, requiring careful attention to occlusions, motion blur, and edge cases. Quality assurance was a top priority—I followed strict labeling guidelines, participated in regular calibration reviews, and maintained over 98% accuracy across validation checks. My contributions helped improve object tracking and scene understanding for real-time decision-making in autonomous navigation.

2020 - 2024

Education

Work History