Jephthah Odei

Key Skills

Software

CVAT

Labelbox

Top Subject Matter

AI Model Training/Evaluation

LLM Prompt Engineering/Quality Assurance

STEM/AI Model Education Data Generation

Top Data Types

Text

Video

Audio

Image

Document

Top Task Types

Prompt Response Writing SFT

Action Recognition

Transcription

Freelancer Overview

Generalist AI Research Expert. Brings 8+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Internal, Proprietary Tooling, and Micro1. Education includes Doctor of Philosophy, Chattered Institute of Leadership, Manpower & Management Development (2026) and Master of Science, Nasarawa State University (2023). AI-training focus includes data types such as Text, Video, and Audio and labeling workflows including Prompt + Response Writing (SFT), Action Recognition, and Transcription.

Labeling Experience

Prompt Engineering/AI Model Trainer

TextPrompt Response Writing SFT

In this role, AI model (LLM) prompts and deliverables were designed and iteratively refined to enhance accuracy and efficiency. Quality assessments were performed on deliverables to identify areas of improvement and maintain high standards. The position involved collaborating dynamically to simulate academic topics and apply prompt engineering expertise. • Created high-performing prompts for LLMs. • Evaluated 30+ model submissions for quality assurance. • Focused on authentic academic/industrial data generation. • Achieved a 95%+ deliverable authenticity rating.

In this role, AI model (LLM) prompts and deliverables were designed and iteratively refined to enhance accuracy and efficiency. Quality assessments were performed on deliverables to identify areas of improvement and maintain high standards. The position involved collaborating dynamically to simulate academic topics and apply prompt engineering expertise. • Created high-performing prompts for LLMs. • Evaluated 30+ model submissions for quality assurance. • Focused on authentic academic/industrial data generation. • Achieved a 95%+ deliverable authenticity rating.

2025 - 2026

Generalist AI Research Expert

TextPrompt Response Writing SFT

This role involved designing complex prompts and analyzing large language model (LLM) behavior to improve model performance. High-quality prompts, rubrics, and style guides were developed to assess and enhance AI model capabilities. The work ensured logical, clear, and accurate data labeling and model evaluation for fine-tuning purposes. • Designed test prompts and responses for LLM assessment. • Authored detailed evaluation frameworks. • Synthesized and interpreted multimodal training data. • Ensured QA adherence in labeling outputs.

This role involved designing complex prompts and analyzing large language model (LLM) behavior to improve model performance. High-quality prompts, rubrics, and style guides were developed to assess and enhance AI model capabilities. The work ensured logical, clear, and accurate data labeling and model evaluation for fine-tuning purposes. • Designed test prompts and responses for LLM assessment. • Authored detailed evaluation frameworks. • Synthesized and interpreted multimodal training data. • Ensured QA adherence in labeling outputs.

2025 - 2025

Conversational AI Model Trainer, Prompt/Response Evaluator

OneformaTextPrompt Response Writing SFT

In this capacity, I authored challenging, domain-specific problem sets and annotations, pushing the boundaries of AI reasoning. I created and refined clear, accurate prompts and responses, and conducted side-by-side evaluations for model improvement. Adversarial testing and detailed feedback played a key role in advancing AI safety and coherence. • Wrote and reviewed high-quality prompts and responses. • Evaluated outputs for factual accuracy and tone. • Provided comprehensive feedback using rubrics. • Conducted red teaming to identify model vulnerabilities.

In this capacity, I authored challenging, domain-specific problem sets and annotations, pushing the boundaries of AI reasoning. I created and refined clear, accurate prompts and responses, and conducted side-by-side evaluations for model improvement. Adversarial testing and detailed feedback played a key role in advancing AI safety and coherence. • Wrote and reviewed high-quality prompts and responses. • Evaluated outputs for factual accuracy and tone. • Provided comprehensive feedback using rubrics. • Conducted red teaming to identify model vulnerabilities.

2024 - 2025

AI Model/LLM Trainer

Micro1TextPrompt Response Writing SFT

As an AI Model/LLM Trainer, I developed domain-specific problem sets and authored ground truth answers for use in AI fine-tuning. This required adherence to strict rubrics and the presentation of detailed, stepwise solutions and explanations. The work was essential in generating high-quality labeled data for STEM training modules within large language models. • Crafted comprehensive problem sets for LLM output. • Created detailed annotations and solutions. • Ensured strict quality rubric compliance. • Delivered content spanning various STEM domains.

As an AI Model/LLM Trainer, I developed domain-specific problem sets and authored ground truth answers for use in AI fine-tuning. This required adherence to strict rubrics and the presentation of detailed, stepwise solutions and explanations. The work was essential in generating high-quality labeled data for STEM training modules within large language models. • Crafted comprehensive problem sets for LLM output. • Created detailed annotations and solutions. • Ensured strict quality rubric compliance. • Delivered content spanning various STEM domains.

2024 - 2025

LLM Video Annotation Specialist

CVATVideoAction Recognition

The position involved detailed video annotation of robot movements, breaking down complex actions into granular notes with accurate timestamps. Each recording was assessed for outcome quality and observations were systematically documented. The data contributed directly to training and improving AI models in visual-motor domains. • Utilized specialized tools for frame-by-frame annotation. • Categorized action outcomes as successful, partial, or failed. • Maintained annotation consistency and detail. • Produced high-quality data for LLM video training.

The position involved detailed video annotation of robot movements, breaking down complex actions into granular notes with accurate timestamps. Each recording was assessed for outcome quality and observations were systematically documented. The data contributed directly to training and improving AI models in visual-motor domains. • Utilized specialized tools for frame-by-frame annotation. • Categorized action outcomes as successful, partial, or failed. • Maintained annotation consistency and detail. • Produced high-quality data for LLM video training.

2023 - 2024

Show more

Education

Work History