Eri Watanabe

Key Skills

Software

Scale AI

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Text

Top Task Types

Audio Recording

Evaluation Rating

Freelancer Overview

I have a deep appreciation for the Japanese language and how words can carry both logic and emotion. Through my background in education and nonprofit work, I’ve learned how language shapes connection — especially when people face delicate or emotional experiences. In AI data training and content evaluation, I focus on helping systems understand those subtleties that make human communication so unique. My work involves Japanese linguistic QA, text review, and contextual labeling, always guided by empathy and precision. Having led projects that support people through sensitive life moments, I’ve developed a strong ability to read between the lines and handle emotional nuance with care. My aim is to help create AI that communicates not only correctly, but also kindly — with warmth, respect, and sincerity. While I can work comfortably in both Japanese and English, my real strength lies in bridging data with human meaning, and bringing a touch of humanity to technology.

Labeling Experience

Japanese Audio Transcription and Text Labeling Project

LabelboxAudioText GenerationEvaluation Rating

I worked on projects involving Japanese audio data, where I transcribed spoken content into accurate written text and performed labeling to capture meaning, tone, and speaker context. The tasks required careful listening, attention to linguistic nuance, and consistency across large datasets. My work focused on improving speech recognition and natural language understanding models by ensuring that each transcription accurately reflected human emotion, context, and intent. Quality assurance was conducted through multi-step review and feedback to maintain high accuracy and cultural relevance.

I worked on projects involving Japanese audio data, where I transcribed spoken content into accurate written text and performed labeling to capture meaning, tone, and speaker context. The tasks required careful listening, attention to linguistic nuance, and consistency across large datasets. My work focused on improving speech recognition and natural language understanding models by ensuring that each transcription accurately reflected human emotion, context, and intent. Quality assurance was conducted through multi-step review and feedback to maintain high accuracy and cultural relevance.

2025

Education

Work History