Henry Muthini

Key Skills

Software

AWS SageMaker

Anno-Mage

Appen

Argilla

Axiom AI

Clickworker

CloudFactory

CrowdFlower

CrowdSource

CVAT

Data Annotation Tech

Dataloop

Datatroniq

Datumbox

Datasaur

Datature

Dataturk

Deep Systems

Diffgram

Doccano

Encord

Figure Eight

Google Cloud Vertex AI

Hasty

HiveMind

Humanatic

iMerit

Img Lab

Kili Technology

Labelbox

LabelImg

Label Studio

LightTag

Lionbridge

Mighty AI

Mindrift

OneForma

OpenCV AI Kit (OAK)

Playment

Prodigy

Redbrick AI

Remotasks

Roboflow

Sama

Scale AI

Sloth

Snorkel AI

SuperAnnotate

Supervisely

Surge AI

Tagtog

Toloka

Telus

Trilldata Technologies

VoTT

V7 Labs

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Action Recognition

Audio Recording

Classification

Question Answering

Text Summarization

Freelancer Overview

I am an experienced AI data labeling specialist with over 3 years of hands-on involvement in training data for text, image, and audio tasks. My work spans across large language model (LLM) evaluation, text classification, sentiment analysis, image tagging, product categorization, and basic speech transcription. I have contributed to projects involving e-commerce platforms, social media content moderation, and multilingual AI systems—particularly in English and Swahili. My strengths lie in accuracy, attention to detail, and the ability to adapt quickly to different labeling tools and workflows. I have worked with platforms like Label Studio, Remotasks, and other custom annotation tools for both structured and unstructured data. My linguistic versatility, combined with real-world experience in customer-facing roles and digital content management, enables me to consistently deliver high-quality annotations for AI training and model fine-tuning.

Labeling Experience

Multilingual LLM Prompt Evaluation & Response Ranking

Label StudioTextText GenerationEvaluation Rating

Evaluated and rated AI-generated responses to prompts in English and Swahili. Tasks included assessing grammar, coherence, factual accuracy, and relevance. Compared multiple outputs for quality ranking, and provided feedback for fine-tuning large language models. Maintained annotation quality above 97% across weekly review audits.

Evaluated and rated AI-generated responses to prompts in English and Swahili. Tasks included assessing grammar, coherence, factual accuracy, and relevance. Compared multiple outputs for quality ranking, and provided feedback for fine-tuning large language models. Maintained annotation quality above 97% across weekly review audits.

2022

E-commerce Product Categorization & Image Tagging

Label StudioTextQuestion AnsweringText Generation

Labeled and categorized thousands of product images and descriptions for an online retail platform. Tasks included assigning correct product types, tagging attributes like color, size, and style, and ensuring consistency across the dataset. Contributed to training recommendation and search algorithms. Followed strict quality guidelines with a 95%+ accuracy rate.

Labeled and categorized thousands of product images and descriptions for an online retail platform. Tasks included assigning correct product types, tagging attributes like color, size, and style, and ensuring consistency across the dataset. Contributed to training recommendation and search algorithms. Followed strict quality guidelines with a 95%+ accuracy rate.

2022 - 2023

English/Swahili Speech Transcription & Speaker Labeling

Scale AIAudioTranslation LocalizationAudio Recording

Transcribed voice clips from English and Swahili speakers, applied timestamps, and identified speaker turns in conversations. Followed detailed style guides and tagging rules to support multilingual speech recognition models. Ensured clarity and formatting accuracy, achieving consistent 95%+ transcription quality.

Transcribed voice clips from English and Swahili speakers, applied timestamps, and identified speaker turns in conversations. Followed detailed style guides and tagging rules to support multilingual speech recognition models. Ensured clarity and formatting accuracy, achieving consistent 95%+ transcription quality.

2021 - 2022

Education

Work History