Khawla Belabadia

Key Skills

Software

Clickworker

CVAT

Data Annotation Tech

Remotasks

Scale AI

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Audio Recording

Classification

Data Collection

Entity Ner Classification

Text Generation

Freelancer Overview

I am a trilingual transcription and data annotation specialist fluent in Arabic, French, and English, with experience delivering high-quality, multilingual audio transcription for clients in France, India, and beyond. My work includes timestamping, speaker labeling, and phonetic transcription using the ARPABET system, applied across business, education, and customer service content. In addition to transcription, I have worked on multi-turn chat evaluation, text classification, and sentiment analysis, supported by my background in managing multilingual communication for brands like Bellincio and BourseAtlas. I’ve also completed extensive hands-on training with CVAT, gaining full familiarity with the platform. My experience includes image classification, bounding box labeling, and video interpolation for human-related content such as fashion items, facial expressions, and actions (walking, sitting, waving). I’ve practiced for 6 months using open datasets like UCF101 and FER-2013, and I’m confident in managing labeling workflows, maintaining annotation consistency, and delivering high-quality training data for AI systems.

Labeling Experience

Multilingual Transcription & AI Chat Evaluation

LighttagTextEntity Ner ClassificationClassification

As a freelance transcriptionist and AI evaluation specialist, I worked on multiple projects involving audio transcription, ARPABET phonetic transcription, text annotation, and AI-generated chat improvement. My tasks included: - Transcribing audio/video in Arabic (native), French, and English with high linguistic precision. - Rewriting or refining weak AI responses to improve natural flow and correctness. - Performing audio segmentation, speaker identification, and time-stamping. - Ensuring high-quality deliverables by adhering to strict QA guidelines and feedback cycles.

As a freelance transcriptionist and AI evaluation specialist, I worked on multiple projects involving audio transcription, ARPABET phonetic transcription, text annotation, and AI-generated chat improvement. My tasks included: - Transcribing audio/video in Arabic (native), French, and English with high linguistic precision. - Rewriting or refining weak AI responses to improve natural flow and correctness. - Performing audio segmentation, speaker identification, and time-stamping. - Ensuring high-quality deliverables by adhering to strict QA guidelines and feedback cycles.

2023 - 2025

Human-Centered Image & Video Annotation (CVAT Training)

CVATVideoBounding BoxClassification

Completed a self-directed annotation training program using CVAT, working with open datasets such as UCF101 (human action recognition), FER2013 (facial expressions), and fashion product images. Tasks included drawing bounding boxes, classifying facial expressions (e.g. happy, sad, angry), tracking human actions (e.g. walking, sitting), and labeling clothing types (e.g. dress, scarf). Developed familiarity with CVAT tools, timeline navigation, interpolation, annotation export formats, and quality control. Labeled and reviewed 80+ samples. Focused on annotation accuracy, consistency, and best practices.

Completed a self-directed annotation training program using CVAT, working with open datasets such as UCF101 (human action recognition), FER2013 (facial expressions), and fashion product images. Tasks included drawing bounding boxes, classifying facial expressions (e.g. happy, sad, angry), tracking human actions (e.g. walking, sitting), and labeling clothing types (e.g. dress, scarf). Developed familiarity with CVAT tools, timeline navigation, interpolation, annotation export formats, and quality control. Labeled and reviewed 80+ samples. Focused on annotation accuracy, consistency, and best practices.

2025 - 2025

Human-Centered Image & Video Annotation (CVAT Training)

CVATImageBounding BoxClassification

Completed a self-directed annotation training program using CVAT, working with open datasets such as UCF101 (human action recognition), FER2013 (facial expressions), and fashion product images. Tasks included drawing bounding boxes, classifying facial expressions (e.g. happy, sad, angry), tracking human actions (e.g. walking, sitting), and labeling clothing types (e.g. dress, scarf). Developed familiarity with CVAT tools, timeline navigation, interpolation, annotation export formats, and quality control. Labeled and reviewed 80+ samples. Focused on annotation accuracy, consistency, and best practices.

Completed a self-directed annotation training program using CVAT, working with open datasets such as UCF101 (human action recognition), FER2013 (facial expressions), and fashion product images. Tasks included drawing bounding boxes, classifying facial expressions (e.g. happy, sad, angry), tracking human actions (e.g. walking, sitting), and labeling clothing types (e.g. dress, scarf). Developed familiarity with CVAT tools, timeline navigation, interpolation, annotation export formats, and quality control. Labeled and reviewed 80+ samples. Focused on annotation accuracy, consistency, and best practices.

2025 - 2025

Multilingual Audio Transcription for Global Clients (EN/FR/AR)

OtherAudioClassificationText Generation

Performed multilingual audio transcription for clients in France, India, and North Africa. Tasks included speaker labeling, timestamping, accurate transcription in French, English, and Moroccan Arabic, and basic emotion/tone classification based on customer service scenarios. Transcripts were delivered manually using formatting guidelines and quality assurance checks, with client feedback loops for continuous improvement. This project helped prepare me for phonetic transcription (ARPABET) and future AI training use cases.

Performed multilingual audio transcription for clients in France, India, and North Africa. Tasks included speaker labeling, timestamping, accurate transcription in French, English, and Moroccan Arabic, and basic emotion/tone classification based on customer service scenarios. Transcripts were delivered manually using formatting guidelines and quality assurance checks, with client feedback loops for continuous improvement. This project helped prepare me for phonetic transcription (ARPABET) and future AI training use cases.

2022 - 2024

Education

Work History