Kazunori Oohashi

Key Skills

Software

Other

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Geospatial Tiled Imagery

Text

Top Task Types

Text Generation

Text Summarization

Translation Localization

Freelancer Overview

I am a native Japanese professional with hands-on experience in AI training and data labeling, particularly in audio transcription, linguistic annotation, and quality control. Over the past year, I have led a local AI project in Japan focused on building a fine-tuned large language model for a multilingual tourism guidance system. In this project, I conducted Japanese and Okinawan dialect speech transcription, annotation for natural conversation, and dataset preparation for LLM fine-tuning. I also developed an automated data validation workflow using GitHub Actions and integrated the Claude Code SDK to perform real-time consistency checks and metadata verification during the annotation process. This framework ensured reproducibility, transparency, and higher annotation accuracy. My background includes extensive work with Whisper-based ASR tools for Japanese transcription, translation between Japanese and English, and text correction using large language models. I am detail-oriented, follow style guides precisely, and maintain high accuracy in multi-speaker and mixed-script transcription. With strong linguistic intuition and technical literacy, I can deliver high-quality datasets that meet AI model training standards.

Labeling Experience

Education

Work History