Jinrun L.

Key Skills

Software

Don't disclose

Other

Top Subject Matter

Multilingual speech data annotation

Transcription Domain Expertise

and quality auditing for AI model training/evaluation

Top Data Types

Audio

Text

Document

Top Task Types

Transcription

Emotion Recognition

Freelancer Overview

Data Annotator — Audio/Text labeling, transcription, verification, and quality inspection (AI projects at ByteDance and . Brings 1+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose and Other. Education includes Master of Arts, Wuhan University of Technology (2026) and Bachelor of Engineering, East China Jiaotong University (2021). AI-training focus includes data types such as Audio and labeling workflows including Transcription and Emotion Recognition.

Labeling Experience

Data Annotator — Audio/Text labeling, transcription, verification, and quality inspection (AI projects at ByteDance and Tencent)

Don't discloseAudioTranscription

Worked as a Data Annotator on multiple AI projects, producing high-accuracy labeled audio and related text artifacts for large language model iteration. Performed transcription, spoken answer verification, and auditory evaluations of user voice feedback to support model optimization and grading. Ensured consistent decisions under noisy, accented, and complex multi-turn speech conditions. • Labeled audio & text data for ByteDance “Creation Bench” with quality inspection • Conducted auditory evaluation and scoring for Tencent user voice feedback to improve intent recognition • Verified spoken answers and performed accuracy assessment/grading annotation for Tencent multi-turn audio tasks • Followed multilingual annotation guidelines to deliver schedule-aligned, high-quality dataset contributions

2026 - Present

Project Experience — Social Media Content Risk Analysis & Judgment (speech emotion + semantic risk assessment)

OtherAudioEmotion Recognition

Developed a content risk discrimination and evaluation approach using mainstream public audio and text to identify and grade spoken/text risk. Applied speech emotion and semantic analysis to support risk detection and disposal suggestion formulation for clustering public opinion. Produced standardized risk reports after processing large-scale cross-platform content. • Crawled, transcribed, sorted, and analyzed 100,000+ cross-platform audio/text items via SQL & Python • Leveraged generative AI tools for semantic and speech emotion analysis to improve risk discrimination efficiency • Performed grading assessment and generated disposal suggestion formulations for risk clusters • Delivered standardized analysis reports and contributed to project selection under an academic innovation fund

2025 - 2025

Education

Work History