Rena Senoman

Key Skills

Software

Toloka

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Top Label Types

Classification

Emotion Recognition

Transcription

Freelancer Overview

I am a freelance audio and speech data annotator with over two years of experience specializing in English transcription, annotation, and linguistic analysis for remote AI and language projects. My work has focused on producing highly accurate transcripts with precise timestamps, detailed speaker labels, and contextual notes for a variety of audio and video content, including interviews, podcasts, technical recordings, and educational lectures. I am skilled at analyzing emotion, tone, and communicative intent, as well as explaining grammar and stylistic choices in clear, accessible language. My background in music training has strengthened my attention to rhythm, emphasis, and subtle shifts in spoken communication, which enhances my annotation quality. I am comfortable working with detailed style guides, tight deadlines, and distributed teams, and I regularly use web-based transcription and annotation tools. I am committed to maintaining high standards of accuracy, confidentiality, and collaboration to support the development of robust AI training data.

Labeling Experience

Call Center Audio Annotation for Sentiment and Speech Analysis

TolokaAudioClassificationEmotion Recognition

Annotated 150 hours of English-language customer service calls from telecommunications clients. Performed verbatim transcription with precise timestamps for each utterance. Conducted speaker diarization to distinguish between agents and customers in overlapping dialogues. Labeled emotions including frustration, satisfaction, anger, and neutrality based on vocal cues like pitch changes, volume spikes, and pauses. Flagged non-speech elements such as background noise or hold music that impacted context. Reviewed annotations in batches of 20 calls per day to ensure 98% inter-annotator agreement on sentiment tags.

Annotated 150 hours of English-language customer service calls from telecommunications clients. Performed verbatim transcription with precise timestamps for each utterance. Conducted speaker diarization to distinguish between agents and customers in overlapping dialogues. Labeled emotions including frustration, satisfaction, anger, and neutrality based on vocal cues like pitch changes, volume spikes, and pauses. Flagged non-speech elements such as background noise or hold music that impacted context. Reviewed annotations in batches of 20 calls per day to ensure 98% inter-annotator agreement on sentiment tags.

2024 - 2024

Education

Work History