Naveen Kharwar

Key Skills

Software

Label Studio

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Bounding Box

Classification

Translation Localization

Freelancer Overview

AI Data Annotation & Hindi Speech Specialist | Audio Labeling • Transcription • Evaluation I am an experienced AI/ML data annotation professional specialized in Hindi speech, audio labeling, clean & structured transcription, and text-to-audio evaluation for training modern LLMs and ASR (Automatic Speech Recognition) models. I have hands-on experience working on multiple AI training and data quality projects that require: - Audio speech segmentation & labeling - Clean, grammar-corrected Hindi transcription - Disfluency removal & formatting - Conversation-level annotation - Speaker tagging & metadata creation - Utterance-level quality evaluation - Text data labeling & fact evaluation Core Skills - Hindi Speech Annotation - Hindi/English Audio Transcription - Audio Labeling & Segmentation - AI Data Evaluation (Voice + Text) - Text Labeling (Intent, Sentiment, QA Judgement) - Linguistic Quality Checks - Data Cleaning & Formatting Annotation Tools: (Hive, Appen, DataForce, Remotasks, Oneforma, Label Studio, etc.) My Unique Strengths: - Bilingual (Hindi–English) - Detail-oriented annotator with strong linguistic understanding - Experience in clean verbatim transcription for AI training - Consistent and reliable delivery - Quick learner for new tools and guidelines My unique strengths include high accuracy, consistent delivery, and the ability to understand and follow complex annotation guidelines with reliability. I focus on quality as required for AI model improvement.

Labeling Experience

Sentiment Analysis_ Practice Project

Label StudioTextClassification

Manually annotated 100+ short social media posts and sentiment tags (positive, negative, neutral). Learned classification workflows and basics of Natural Language Processing (NLP).

Manually annotated 100+ short social media posts and sentiment tags (positive, negative, neutral). Learned classification workflows and basics of Natural Language Processing (NLP).

2024

Image Labeling - Demo Project

Label StudioImageBounding BoxClassification

Labeled over 50+ images using bounding boxes to detect objects like car, people, and sigs.Practised exporting datasets in COCO format. Focused on accuracy, consistency, and clean annotations suitable for AI model training.

Labeled over 50+ images using bounding boxes to detect objects like car, people, and sigs.Practised exporting datasets in COCO format. Focused on accuracy, consistency, and clean annotations suitable for AI model training.

2024

Education

Work History