Chike Iroham

Key Skills

Software

AWS SageMaker

Appen

Argilla

CVAT

Hasty

Labelbox

LabelImg

Label Studio

Prodigy

Roboflow

Scale AI

SuperAnnotate

Supervisely

V7 Labs

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Audio Recording

Bounding Box

Computer Programming Coding

Data Collection

Entity Ner Classification

Freelancer Overview

I am a data labeling and AI training data specialist with over five years of hands-on experience creating high-accuracy datasets for advanced AI models. My core expertise lies in computer vision for autonomous vehicles, where I have meticulously annotated 2D images, video sequences, and 3D LiDAR point clouds to define critical elements like drivable paths, traffic actors, and obstacles. I am also proficient in LLM evaluation and multilingual text generation, specializing in refining model outputs for safety, accuracy, and linguistic fluency in both English and Spanish. What sets me apart is my dual-domain proficiency, combining deep technical annotation skills with nuanced linguistic analysis. I excel at navigating complex annotation guidelines, implementing rigorous quality control protocols, and adapting quickly to new tooling—from platforms like Labelbox and Scale AI to open-source tools like CVAT. My work directly contributes to both the perceptual systems of self-driving cars and the conversational reliability of large language models.

Labeling Experience

Senior Annotator for Autonomous Vehicle Perception System

Scale AIVideoBounding BoxPolygon

Led annotation for a large-scale dataset to train the perception stack of a Level 4 autonomous vehicle. Tasks included precise 3D cuboid placement on LiDAR point clouds for vehicles/pedestrians, semantic segmentation of drivable surfaces in video sequences, and polygonal annotation for traffic signs. Managed a subset of ~50,000 frames, adhering to strict OEM guidelines and maintaining a >99% QA pass rate through iterative review cycles.

Led annotation for a large-scale dataset to train the perception stack of a Level 4 autonomous vehicle. Tasks included precise 3D cuboid placement on LiDAR point clouds for vehicles/pedestrians, semantic segmentation of drivable surfaces in video sequences, and polygonal annotation for traffic signs. Managed a subset of ~50,000 frames, adhering to strict OEM guidelines and maintaining a >99% QA pass rate through iterative review cycles.

2022 - 2024

Education

Work History