John Maubi

Key Skills

Software

Clickworker

Data Annotation Tech

iMerit

Labelbox

Label Studio

Mindrift

SuperAnnotate

Telus

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Computer Code Programming

Document

Image

Text

Video

Top Task Types

Audio Recording

Bounding Box

Classification

Computer Programming/Coding

Data Collection

Entity (NER) Classification

Evaluation/Rating

Prompt + Response Writing (SFT)

Text Generation

Text Summarization

Transcription

Translation/Localization

Freelancer Overview

I am an experienced data operations and AI training data specialist with a strong background in data labeling, annotation, and quality assurance for international projects. My work includes processing and validating large-scale datasets with over 98% accuracy, transcribing and annotating audio content in both English and Swahili, and improving dataset quality through rigorous quality checks. I am skilled in tools such as MySQL, PostgreSQL, Microsoft Office, Google Workspace, Slack, Teams, and Zoom, and have supported projects in domains like healthcare (EMR systems), linguistic annotation, and general AI data operations. My ability to collaborate with global teams, optimize data workflows, and maintain high standards of accuracy and consistency makes me confident in delivering reliable, high-quality training data for AI and machine learning applications.

Labeling Experience

Large-Scale Multimodal AI Training Data Curation

TelusImageEntity Ner ClassificationClassification

Led end-to-end annotation across 1,500+ hours of video, image, audio, and text datasets for production LLMs and computer vision models at TELUS International and RWS Moravia. Specialized in complex video annotation including temporal segmentation, object tracking, action recognition, and scene classification with 98%+ accuracy. Conducted extensive LLM evaluation work—refining prompts, assessing model reasoning quality, and testing contextual accuracy across thousands of responses. Performed quality assurance and peer review identifying systematic errors that improved project accuracy by 15-20%. Worked on multilingual projects including 500+ hours of Swahili audio transcription, applying linguistic expertise to low-resource language datasets. Collaborated directly with ML engineering teams via GitHub and Slack to optimize annotation schemas, debug pipeline issues, and ensure dataset quality standards aligned with model perform

Led end-to-end annotation across 1,500+ hours of video, image, audio, and text datasets for production LLMs and computer vision models at TELUS International and RWS Moravia. Specialized in complex video annotation including temporal segmentation, object tracking, action recognition, and scene classification with 98%+ accuracy. Conducted extensive LLM evaluation work—refining prompts, assessing model reasoning quality, and testing contextual accuracy across thousands of responses. Performed quality assurance and peer review identifying systematic errors that improved project accuracy by 15-20%. Worked on multilingual projects including 500+ hours of Swahili audio transcription, applying linguistic expertise to low-resource language datasets. Collaborated directly with ML engineering teams via GitHub and Slack to optimize annotation schemas, debug pipeline issues, and ensure dataset quality standards aligned with model perform

2023 - 2024

AI Data Operations Specialist, TELUS International

TelusTextClassification

As an AI Data Operations Specialist at TELUS International, I processed and validated large-scale training datasets for NLP models. I performed various text annotation tasks such as text classification, sentiment analysis, intent labeling, and entity recognition for machine learning applications. I applied complex annotation guidelines and quality control procedures, evaluated and ranked AI model outputs utilizing RLHF methodologies, and maintained exceptional accuracy rates. • Conducted model evaluation and LLM response assessment using preference ranking and instruction tuning tasks • Contributed to process improvement by reducing quality assurance rework 20% • Collaborated with international teams to maintain high annotation standards • Used web-based dashboard annotation tools similar to Labelbox and Telus platforms

As an AI Data Operations Specialist at TELUS International, I processed and validated large-scale training datasets for NLP models. I performed various text annotation tasks such as text classification, sentiment analysis, intent labeling, and entity recognition for machine learning applications. I applied complex annotation guidelines and quality control procedures, evaluated and ranked AI model outputs utilizing RLHF methodologies, and maintained exceptional accuracy rates. • Conducted model evaluation and LLM response assessment using preference ranking and instruction tuning tasks • Contributed to process improvement by reducing quality assurance rework 20% • Collaborated with international teams to maintain high annotation standards • Used web-based dashboard annotation tools similar to Labelbox and Telus platforms

2023

Data Quality Specialist, RWS Moravia

LabelboxTextEvaluation Rating

As a Data Quality Specialist at RWS Moravia, I executed comprehensive data validation, entry, and annotation audits for multiple AI training projects. I reviewed and audited annotated data outputs, improving overall data accuracy through error identification and edge case analysis. I collaborated with international teams to document annotation protocols, optimize evaluation workflows, and maintain quality standards. • Enhanced data quality by 15% through effective auditing and protocol optimization • Resolved complex data quality challenges in ambiguous annotation scenarios • Applied analytical thinking for comprehensive data review • Utilized web-based dashboards akin to Labelbox for annotation review and audit tasks

As a Data Quality Specialist at RWS Moravia, I executed comprehensive data validation, entry, and annotation audits for multiple AI training projects. I reviewed and audited annotated data outputs, improving overall data accuracy through error identification and edge case analysis. I collaborated with international teams to document annotation protocols, optimize evaluation workflows, and maintain quality standards. • Enhanced data quality by 15% through effective auditing and protocol optimization • Resolved complex data quality challenges in ambiguous annotation scenarios • Applied analytical thinking for comprehensive data review • Utilized web-based dashboards akin to Labelbox for annotation review and audit tasks

2022

Language Specialist | Audio Annotation Expert, Your Personal AI

LabelboxAudioClassification

As a Language Specialist and Audio Annotation Expert for Your Personal AI, I processed and transcribed over 500 hours of audio for AI training datasets. I performed audio classification, speaker identification, and intent labeling to enable accurate speech recognition model development. I ensured bilingual data accuracy (English/Swahili), contextual appropriateness, and improved dataset quality through rigorous validation. • Enhanced audio dataset quality by 25% through systematic review • Applied speaker identification and intent classification for high-fidelity labeling • Utilized advanced audio annotation tools integrated in web platforms (Labelbox-style) • Ensured cross-cultural and linguistic accuracy for NLP applications

As a Language Specialist and Audio Annotation Expert for Your Personal AI, I processed and transcribed over 500 hours of audio for AI training datasets. I performed audio classification, speaker identification, and intent labeling to enable accurate speech recognition model development. I ensured bilingual data accuracy (English/Swahili), contextual appropriateness, and improved dataset quality through rigorous validation. • Enhanced audio dataset quality by 25% through systematic review • Applied speaker identification and intent classification for high-fidelity labeling • Utilized advanced audio annotation tools integrated in web platforms (Labelbox-style) • Ensured cross-cultural and linguistic accuracy for NLP applications

2025 - 2025

Education

Work History