Joshua Mwinamu

Key Skills

Software

CVAT

Other

Top Subject Matter

Natural Language Processing

Computer Vision

Top Data Types

Text

Image

Top Task Types

Segmentation

RLHF

Freelancer Overview

AI Content Evaluator / Freelance / independent contractor. Core strengths include N, A, and CVAT. Education includes Bachelor of Science, University of Nairobi (2025). AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation, Rating, and Segmentation.

Labeling Experience

AI Content Evaluator / Freelance / independent contractor

Text

Oversaw detailed rubric-based evaluations of AI-generated textual content to enhance model safety and accuracy. Utilized systematic comparison and ranking of outputs to refine AI helpfulness. Proactively identified and flagged unsafe or problematic content to reinforce quality standards. • Improved accuracy scores of AI models by 15% through diligent textual content assessment. • Enhanced helpfulness ratings by 20% via qualitative comparison of model outputs. • Ensured safety compliance by detecting and reporting over 50 unsafe content instances. • Independently executed content evaluation as a freelance contractor.

Oversaw detailed rubric-based evaluations of AI-generated textual content to enhance model safety and accuracy. Utilized systematic comparison and ranking of outputs to refine AI helpfulness. Proactively identified and flagged unsafe or problematic content to reinforce quality standards. • Improved accuracy scores of AI models by 15% through diligent textual content assessment. • Enhanced helpfulness ratings by 20% via qualitative comparison of model outputs. • Ensured safety compliance by detecting and reporting over 50 unsafe content instances. • Independently executed content evaluation as a freelance contractor.

2025 - Present

Linguistic Data Specialist & Quality Assurance Lead

OtherTextRLHF

Directed the annotation and evaluation of linguistic datasets to support reinforcement learning from human feedback (RLHF) for large language models. Advanced beyond basic labeling to complex linguistic and behavioral task design for AI intent recognition. Maintained high-quality standards for text data used in model fine-tuning. • Employed RLHF pipelines using JSON format for large language model improvement. • Managed data annotation standards and QA for NLP projects. • Consistently delivered high-precision RLHF tasks in challenging environments. • Provided leadership in multimodal dataset management and linguistic error classification.

Directed the annotation and evaluation of linguistic datasets to support reinforcement learning from human feedback (RLHF) for large language models. Advanced beyond basic labeling to complex linguistic and behavioral task design for AI intent recognition. Maintained high-quality standards for text data used in model fine-tuning. • Employed RLHF pipelines using JSON format for large language model improvement. • Managed data annotation standards and QA for NLP projects. • Consistently delivered high-precision RLHF tasks in challenging environments. • Provided leadership in multimodal dataset management and linguistic error classification.

2024 - 2024

Linguistic Data Specialist & Quality Assurance Lead

CVATImageSegmentation

Led the development and annotation of high-fidelity datasets for computer vision models, emphasizing nuanced human action and intent recognition. Coordinated multimodal data projects integrating complex behavioral and linguistic analysis. Ensured the creation of precisely labeled image data for advanced AI training. • Applied segmentation techniques in CVAT for detailed computer vision annotation. • Developed and QA'd datasets for computer vision and language models. • Maintained 98% accuracy in image-related annotation tasks. • Utilized Python scripting and SQL for dataset handling and quality assurance.

Led the development and annotation of high-fidelity datasets for computer vision models, emphasizing nuanced human action and intent recognition. Coordinated multimodal data projects integrating complex behavioral and linguistic analysis. Ensured the creation of precisely labeled image data for advanced AI training. • Applied segmentation techniques in CVAT for detailed computer vision annotation. • Developed and QA'd datasets for computer vision and language models. • Maintained 98% accuracy in image-related annotation tasks. • Utilized Python scripting and SQL for dataset handling and quality assurance.

2024 - 2024

Education

Work History