Harsh Dharaiya

Key Skills

Software

Scale AI

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Text

Top Label Types

Evaluation Rating

Audio Recording

Text Generation

Freelancer Overview

I am a multilingual storyteller and voice-over artist with extensive experience in creating high-quality audio data for AI training and natural language processing applications. I have recorded over 100 hours of voice content for platforms such as Outlier.com, where I contributed 105+ audio recordings specifically for AI voice model training. My background includes dubbing, scriptwriting, translation, and transcription in Gujarati, Hindi, and English, ensuring precise and culturally relevant data annotation. I am skilled with tools like Audacity, Adobe Audition, Premiere Pro, and have a strong technical setup for home recording. My experience spans EdTech, podcasts, and creative media, making me adept at producing, labeling, and quality-checking diverse datasets for AI and machine learning projects.

Labeling Experience

Generalist

Internal Proprietary ToolingTextText Generation

I am currently working on the Aether project through Outlier, contributing to AI training in handwritten text recognition and script validation. My role involves reviewing handwritten content on multimango.com and verifying the accuracy of AI-generated transcriptions. I carefully read and interpret handwritten material, compare it against AI-prepared scripts, and identify discrepancies in spelling, structure, and contextual meaning. I ensure linguistic accuracy, semantic correctness, and alignment between the original handwritten input and the AI output. My work supports the improvement of handwriting recognition models and language processing systems by providing precise human validation and corrective feedback.

I am currently working on the Aether project through Outlier, contributing to AI training in handwritten text recognition and script validation. My role involves reviewing handwritten content on multimango.com and verifying the accuracy of AI-generated transcriptions. I carefully read and interpret handwritten material, compare it against AI-prepared scripts, and identify discrepancies in spelling, structure, and contextual meaning. I ensure linguistic accuracy, semantic correctness, and alignment between the original handwritten input and the AI output. My work supports the improvement of handwriting recognition models and language processing systems by providing precise human validation and corrective feedback.

2025

Voice Data Annotator for AI Speech Models

Scale AIAudioEvaluation RatingAudio Recording

I worked with Scale AI as a Speech Data Contributor, providing studio-quality voice recordings for AI model training. I received scripted content, carefully reviewed and evaluated it for context and linguistic accuracy, emotion, grammar and then recorded clear, natural, and consistent voice samples. I maintained precise pronunciation, controlled pacing, tonal consistency, and clean audio standards to meet ASR (Automatic Speech Recognition) and TTS (Text-to-Speech) data requirements. My recordings were delivered as structured speech data to support supervised machine learning workflows.

I worked with Scale AI as a Speech Data Contributor, providing studio-quality voice recordings for AI model training. I received scripted content, carefully reviewed and evaluated it for context and linguistic accuracy, emotion, grammar and then recorded clear, natural, and consistent voice samples. I maintained precise pronunciation, controlled pacing, tonal consistency, and clean audio standards to meet ASR (Automatic Speech Recognition) and TTS (Text-to-Speech) data requirements. My recordings were delivered as structured speech data to support supervised machine learning workflows.

2025 - 2025

Education

Work History