Collins Napoleon

Key Skills

Software

Scale AI

Top Subject Matter

Vision-Language AI Training

Multimodal AI Model Training

Language Model Training – LLM/RLHF

Top Data Types

Image

Video

Text

Top Task Types

Classification

Action Recognition

RLHF

Freelancer Overview

AI Training Specialist – Image Annotation and Description. Brings 11+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Scale AI. Education includes Master of Science, York St John University (2026) and Bachelor of Engineering, University of Port Harcourt (2012). AI-training focus includes data types such as Image, Video, and Text and labeling workflows including Classification, Action Recognition, and RLHF.

Labeling Experience

AI Training Specialist – RLHF & Response Evaluation

Scale AITextRLHF

I evaluated and refined large language model outputs for clarity, coherence, and instruction-following using RLHF methodologies on the Aether project. Tasked with reviewing AI-generated text, I flagged errors, rewrote responses, and assessed outputs against multi-criteria rubrics. My role emphasized factual accuracy, logical consistency, and high-quality language use. • Conducted prompt writing and response ranking to enhance model performance • Identified and addressed hallucinations and factual inaccuracies • Followed detailed style guides and rubric-based evaluation protocols • Maintained consistent high quality and pass rates in annotation tasks

I evaluated and refined large language model outputs for clarity, coherence, and instruction-following using RLHF methodologies on the Aether project. Tasked with reviewing AI-generated text, I flagged errors, rewrote responses, and assessed outputs against multi-criteria rubrics. My role emphasized factual accuracy, logical consistency, and high-quality language use. • Conducted prompt writing and response ranking to enhance model performance • Identified and addressed hallucinations and factual inaccuracies • Followed detailed style guides and rubric-based evaluation protocols • Maintained consistent high quality and pass rates in annotation tasks

2025 - 2025

AI Training Specialist – Video Annotation

Scale AIVideoAction Recognition

On the Aether project, I performed detailed frame-by-frame annotation of video content to enhance training data for multimodal AI models. This involved labeling actions, objects, and scene transitions, with a focus on accuracy and adherence to evolving task rubrics. I evaluated and corrected model interpretations to ensure output quality and minimize errors. • Provided objective frame-level descriptions and action tagging across video sequences • Adhered to strict annotation standards and updated guidelines • Flagged logical inconsistencies and hallucinations in model-generated content • Supported large-scale, remote video annotation and content labeling tasks

On the Aether project, I performed detailed frame-by-frame annotation of video content to enhance training data for multimodal AI models. This involved labeling actions, objects, and scene transitions, with a focus on accuracy and adherence to evolving task rubrics. I evaluated and corrected model interpretations to ensure output quality and minimize errors. • Provided objective frame-level descriptions and action tagging across video sequences • Adhered to strict annotation standards and updated guidelines • Flagged logical inconsistencies and hallucinations in model-generated content • Supported large-scale, remote video annotation and content labeling tasks

2025 - 2025

AI Training Specialist – Image Annotation and Description

Scale AIImageClassification

As an AI Training Specialist on the Aether project via Outlier and Scale AI, I created detailed image descriptions and annotations for vision-language AI model training. My work involved crafting precise visual captions, object recognition, and scene analysis while closely adhering to annotation guidelines. I maintained high consistency and accuracy when labeling and reviewing visual data outputs. • Generated high-quality image captions and alt-text for diverse visual content • Followed strict project guidelines for annotation consistency and accuracy • Identified and flagged visual misinterpretations and errors in model outputs • Employed strong attention to detail and descriptive writing skills to improve data quality

As an AI Training Specialist on the Aether project via Outlier and Scale AI, I created detailed image descriptions and annotations for vision-language AI model training. My work involved crafting precise visual captions, object recognition, and scene analysis while closely adhering to annotation guidelines. I maintained high consistency and accuracy when labeling and reviewing visual data outputs. • Generated high-quality image captions and alt-text for diverse visual content • Followed strict project guidelines for annotation consistency and accuracy • Identified and flagged visual misinterpretations and errors in model outputs • Employed strong attention to detail and descriptive writing skills to improve data quality

2025 - 2025

Education

Work History