Triumph

Key Skills

Software

Other

Remotasks

Appen

Top Subject Matter

LLM Outputs

AI Safety

Rlhf Domain Expertise

Top Data Types

Text

Image

Audio

Document

Top Task Types

Red Teaming

Entity (NER) Classification

Bounding Box

Transcription

Classification

Freelancer Overview

AI Data Trainer & LLM Evaluation Lead. Core strengths include Internal, Proprietary Tooling, and Other. Education includes Master of Science, University of Lagos (2024) and Bachelor of Science, University of Uyo (2021). AI-training focus includes data types such as Text, Image, and Audio and labeling workflows including Evaluation, Rating, and Red Teaming.

Labeling Experience

AI Data Trainer & LLM Evaluation Lead

Text

Led a distributed team of annotators to evaluate and fine-tune large language model outputs for safety, factual accuracy, and instruction adherence. Developed and maintained RLHF annotation guidelines and rubrics used in multiple LLM training projects. Oversaw advanced red-teaming and adversarial prompt testing to expose model failure modes and biases. • Curated and annotated a benchmark dataset of 5,000+ prompt-response pairs for creative and factual Q&A. • Delivered weekly quality reports with inter-annotator agreement analyses, achieving a 98% consistency score. • Designed multi-dimensional scoring rubrics for comprehensive LLM output evaluation. • Used internal/proprietary tools and collaborative platforms for annotation and reporting.

Led a distributed team of annotators to evaluate and fine-tune large language model outputs for safety, factual accuracy, and instruction adherence. Developed and maintained RLHF annotation guidelines and rubrics used in multiple LLM training projects. Oversaw advanced red-teaming and adversarial prompt testing to expose model failure modes and biases. • Curated and annotated a benchmark dataset of 5,000+ prompt-response pairs for creative and factual Q&A. • Delivered weekly quality reports with inter-annotator agreement analyses, achieving a 98% consistency score. • Designed multi-dimensional scoring rubrics for comprehensive LLM output evaluation. • Used internal/proprietary tools and collaborative platforms for annotation and reporting.

2025 - 2026

Image/Video Data Annotation Specialist

RemotasksImageBounding Box

Annotated thousands of images and video frames using bounding boxes and polygons for autonomous vehicle and object detection models. Managed labeling workflows to ensure high-quality training data for ML engineering teams. Participated in regular quality control reviews and updated annotation protocols as needed. • Tasks included vehicle, pedestrian, and road sign marking on street-level images. • Used proprietary and open-source annotation software (e.g., CVAT, Labelbox). • Developed and applied strict annotation quality guidelines. • Oversaw multilingual audio transcription and labeling for related datasets.

Annotated thousands of images and video frames using bounding boxes and polygons for autonomous vehicle and object detection models. Managed labeling workflows to ensure high-quality training data for ML engineering teams. Participated in regular quality control reviews and updated annotation protocols as needed. • Tasks included vehicle, pedestrian, and road sign marking on street-level images. • Used proprietary and open-source annotation software (e.g., CVAT, Labelbox). • Developed and applied strict annotation quality guidelines. • Oversaw multilingual audio transcription and labeling for related datasets.

2023 - 2025

Multilingual NER Annotation Specialist

OtherTextEntity Ner Classification

Annotated multilingual named entities in over 10,000 sentences for English and Pidgin as part of an NLP research initiative for low-resource African languages. Ensured high-quality labeling for model training and evaluation. Collaborated with researchers to calibrate guidelines for linguistic and cultural accuracy. • Labeled person, location, organization, and miscellaneous entities. • Followed comprehensive NER annotation guidelines. • Used proprietary or open-source annotation tools (e.g., Doccano, Labelbox). • Supported language-specific disambiguation for higher dataset integrity.

Annotated multilingual named entities in over 10,000 sentences for English and Pidgin as part of an NLP research initiative for low-resource African languages. Ensured high-quality labeling for model training and evaluation. Collaborated with researchers to calibrate guidelines for linguistic and cultural accuracy. • Labeled person, location, organization, and miscellaneous entities. • Followed comprehensive NER annotation guidelines. • Used proprietary or open-source annotation tools (e.g., Doccano, Labelbox). • Supported language-specific disambiguation for higher dataset integrity.

2023 - 2025

Red-Team Adversarial Prompt Dataset Annotator

TextRed Teaming

Executed 3,000+ adversarial prompts that targeted jailbreaking, bias elicitation, and hallucination tests in LLMs to assess safety vulnerabilities. Categorized model failure modes and produced a taxonomy to improve model guardrails and content filters. Collaborated cross-functionally to deliver actionable insights for engineering teams. • Labeling focused on prompt-based attacks and structured output categorization. • Helped define safety evaluation benchmarks for production-grade LLMs. • Delivered structured reports and recommendations for enhanced filtering. • Utilized internal/proprietary tools for red-teaming execution and results documentation.

Executed 3,000+ adversarial prompts that targeted jailbreaking, bias elicitation, and hallucination tests in LLMs to assess safety vulnerabilities. Categorized model failure modes and produced a taxonomy to improve model guardrails and content filters. Collaborated cross-functionally to deliver actionable insights for engineering teams. • Labeling focused on prompt-based attacks and structured output categorization. • Helped define safety evaluation benchmarks for production-grade LLMs. • Delivered structured reports and recommendations for enhanced filtering. • Utilized internal/proprietary tools for red-teaming execution and results documentation.

2024 - 2024

Search Quality Evaluator & AI Data Contributor

AppenText

Evaluated search engine results for relevance, accuracy, and intent match across text, image, and local queries as part of AI data labeling workflows. Labeled image, text, and ad content for algorithm training using strict annotation guidelines. Maintained high consistency and quality benchmarks set by the project. • Worked on content moderation and linguistic evaluation tasks. • Performed labeling for ad relevance and search quality rating. • Delivered feedback on linguistic appropriateness for English-language AI products. • Utilized Appen platform and internal tools for annotation.

Evaluated search engine results for relevance, accuracy, and intent match across text, image, and local queries as part of AI data labeling workflows. Labeled image, text, and ad content for algorithm training using strict annotation guidelines. Maintained high consistency and quality benchmarks set by the project. • Worked on content moderation and linguistic evaluation tasks. • Performed labeling for ad relevance and search quality rating. • Delivered feedback on linguistic appropriateness for English-language AI products. • Utilized Appen platform and internal tools for annotation.

2022 - 2023

Show more

Education

Work History