Goodness Madumere

Key Skills

Software

Appen

LabelImg

Mindrift

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Label Types

Classification

Translation Localization

Freelancer Overview

I am an experienced AI model tester and evaluator with a strong background in data annotation, labeling, and training data quality for NLP applications. My expertise includes evaluating AI-generated text for accuracy, tone, bias, cultural sensitivity, and factual correctness across multiple languages, including English, French, Spanish, Portuguese, and Pidgin. I am skilled in prompt engineering, response ranking, RLHF workflows, and applying complex annotation guidelines to deliver structured, high-quality feedback. I have hands-on experience with tools such as ChatGPT, HeyGen, VEED.io, and Canva, and am adept at comparative analysis, content moderation, and multilingual content review. My work spans diverse domains including content localization, social media, and cross-cultural communication, and I am committed to ensuring data quality and model performance through meticulous evaluation and documentation. I am available for full-time remote roles and flexible across time zones.

Labeling Experience

Audio data labeling

AppenAudioText Generation

Audio Data Labeling Specialist — Freelance 2024 – Present Transcribed and labeled audio recordings across multiple accents and dialects, ensuring accurate text alignment with spoken content for speech recognition model training. Annotated audio files by tagging speaker identity, emotion, intent, background noise levels, and speech clarity to support the development of robust voice AI systems. Applied phonetic and linguistic knowledge to label code-switching and multilingual audio content in English, French, Spanish, and Pidgin — improving model accuracy for diverse speaker demographics. Reviewed and quality-checked audio transcriptions submitted by other annotators, flagging inconsistencies and ensuring adherence to dataset style guides and labeling taxonomies. Categorized audio segments by content type including conversational speech, commands, narration, and emotional tone to build varied and balanced training datasets. Maintained high accuracy and consistency across large-volume au

Audio Data Labeling Specialist — Freelance 2024 – Present Transcribed and labeled audio recordings across multiple accents and dialects, ensuring accurate text alignment with spoken content for speech recognition model training. Annotated audio files by tagging speaker identity, emotion, intent, background noise levels, and speech clarity to support the development of robust voice AI systems. Applied phonetic and linguistic knowledge to label code-switching and multilingual audio content in English, French, Spanish, and Pidgin — improving model accuracy for diverse speaker demographics. Reviewed and quality-checked audio transcriptions submitted by other annotators, flagging inconsistencies and ensuring adherence to dataset style guides and labeling taxonomies. Categorized audio segments by content type including conversational speech, commands, narration, and emotional tone to build varied and balanced training datasets. Maintained high accuracy and consistency across large-volume au

2024

Data Labeling & Annotation Specialist — Freelance / Self-Initiated

AppenTextClassification

Performed multilingual text annotation, sentiment labeling, intent classification, and topic tagging for natural language processing model training. Applied named entity recognition (NER) for tagging persons, organizations, dates, locations, and events in diverse language corpora. Completed structured image descriptions and visual content annotation following detailed dataset style guides. • Labeled content quality in terms of fluency, coherence, accuracy, tone, and cultural appropriateness. • Maintained high annotation consistency scores with rigorous quality assurance. • Adopted annotation platform standards comparable to Scale AI, Appen, Lionbridge, and Remotasks. • Categorized and tagged content for social media, marketing, and editorial datasets.

Performed multilingual text annotation, sentiment labeling, intent classification, and topic tagging for natural language processing model training. Applied named entity recognition (NER) for tagging persons, organizations, dates, locations, and events in diverse language corpora. Completed structured image descriptions and visual content annotation following detailed dataset style guides. • Labeled content quality in terms of fluency, coherence, accuracy, tone, and cultural appropriateness. • Maintained high annotation consistency scores with rigorous quality assurance. • Adopted annotation platform standards comparable to Scale AI, Appen, Lionbridge, and Remotasks. • Categorized and tagged content for social media, marketing, and editorial datasets.

2024

Content Localisation Specialist (Multilingual QA) — Freelance

LionbridgeTextTranslation Localization

Reviewed and validated translated and localized content following structured QA checklists based on data labeling principles. Categorized content by type, region, language register, and audience for use in AI training pipelines. Ensured entity-level consistency and adherence to NER best practices across multilingual datasets. • Collaborated to maintain labeling standards across large content batches. • Cross-referenced terminology databases for multilingual accuracy. • Validated localized text for classification and labeling consistency. • Applied quality control workflows to review dataset integrity.

Reviewed and validated translated and localized content following structured QA checklists based on data labeling principles. Categorized content by type, region, language register, and audience for use in AI training pipelines. Ensured entity-level consistency and adherence to NER best practices across multilingual datasets. • Collaborated to maintain labeling standards across large content batches. • Cross-referenced terminology databases for multilingual accuracy. • Validated localized text for classification and labeling consistency. • Applied quality control workflows to review dataset integrity.

2025 - 2025

Education

Work History