Nicholaus Mlangwa

Key Skills

Software

OneForma

Toloka

Other

Top Subject Matter

Swahili Linguistic QA & Text Evaluation

Text and Image Data Annotation for AI

Voice Data Collection and Validation

Top Data Types

Text

Audio

Image

Top Task Types

Data Collection

Audio Recording

Freelancer Overview

Swahili QA & Linguistic Data Expert (Freelance) – OneForma/Centific (Sagittarius Project). Brings 6+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include OneForma, Toloka, and Other. Education includes Bachelor of Arts, Institute of Rural Development Planning (2013) and Advanced Certificate of Secondary Education, Mzumbe Secondary School (2009). AI-training focus includes data types such as Text and Audio and labeling workflows including Evaluation, Rating, and Data Collection.

Labeling Experience

Voice Data Collector – Silencio Voice AI

OtherAudioAudio Recording

For Silencio Voice AI, I recorded and validated Swahili voice datasets ensuring phonetic accuracy and regional dialect fidelity. The tasks involved careful recording processes and validation of audio for model training. My focus was to ensure all audio matched native Tanzanian Swahili features for voice AI purposes. • Recorded Swahili audio samples for dataset creation. • Validated phonetic properties in collected audio. • Ensured alignment with Tanzanian dialectal norms. • Supported model development for speech AI systems.

For Silencio Voice AI, I recorded and validated Swahili voice datasets ensuring phonetic accuracy and regional dialect fidelity. The tasks involved careful recording processes and validation of audio for model training. My focus was to ensure all audio matched native Tanzanian Swahili features for voice AI purposes. • Recorded Swahili audio samples for dataset creation. • Validated phonetic properties in collected audio. • Ensured alignment with Tanzanian dialectal norms. • Supported model development for speech AI systems.

2024 - Present

Data Annotator – Toloka

TolokaTextData Collection

On Toloka, I performed data annotation and classification tasks focusing on text and images to refine machine learning algorithms. My work included classifying content and annotating datasets for supervised learning. I contributed to the improvement of both training data and algorithmic models via detailed labeling efforts. • Executed data annotation tasks (text/image classification). • Contributed labeled data to AI model training pipelines. • Ensured quality control of submitted labels. • Followed project guidelines for consistency and accuracy.

On Toloka, I performed data annotation and classification tasks focusing on text and images to refine machine learning algorithms. My work included classifying content and annotating datasets for supervised learning. I contributed to the improvement of both training data and algorithmic models via detailed labeling efforts. • Executed data annotation tasks (text/image classification). • Contributed labeled data to AI model training pipelines. • Ensured quality control of submitted labels. • Followed project guidelines for consistency and accuracy.

2024 - Present

Swahili QA & Linguistic Data Expert (Freelance) – OneForma/Centific (Sagittarius Project)

OneformaText

As a Swahili Quality Assurance Specialist, I evaluated and validated linguistic data to ensure high-quality Swahili output for AI models. My role concerned refining grammar, cultural nuances, and factual correctness in large language models using tools like ChatGPT, Gemini, and Claude. I also worked on active training of models and validated outputs for accuracy and relevancy. • Evaluated machine-generated Swahili text for linguistic accuracy. • Conducted prompt engineering for LLM performance improvements. • Participated in QA processes for Swahili data with project teams. • Applied culturally informed assessments to AI outputs.

As a Swahili Quality Assurance Specialist, I evaluated and validated linguistic data to ensure high-quality Swahili output for AI models. My role concerned refining grammar, cultural nuances, and factual correctness in large language models using tools like ChatGPT, Gemini, and Claude. I also worked on active training of models and validated outputs for accuracy and relevancy. • Evaluated machine-generated Swahili text for linguistic accuracy. • Conducted prompt engineering for LLM performance improvements. • Participated in QA processes for Swahili data with project teams. • Applied culturally informed assessments to AI outputs.

2024 - Present

Education

Work History