Kawtar Latifi

Key Skills

Software

CVAT

Img Lab

Labelbox

Label Studio

Remotasks

Scale AI

SuperAnnotate

Top Subject Matter

LLM

TA

HR

Top Data Types

Audio

Image

Video

Top Task Types

Audio Recording

Evaluation Rating

Prompt Response Writing SFT

Question Answering

Freelancer Overview

I have extensive experience in data labeling and AI training data, primarily through my work at Remotask and Outlier AI. At Remotask, I honed my skills in accurately annotating various data types, including images, text, and audio, ensuring high-quality training datasets for machine learning models. This role required meticulous attention to detail, a deep understanding of labeling guidelines, and the ability to adapt to different project requirements swiftly. At Outlier AI, I further advanced my expertise by working on complex data labeling projects that involved identifying and tagging nuanced data patterns. My key responsibilities included managing large datasets, performing quality assurance checks, and collaborating with AI researchers to optimize the training process. My ability to quickly grasp project specifics, coupled with my dedication to maintaining data integrity, sets me apart in the field. These experiences have equipped me with a robust skill set in data labeling, critical for developing accurate and efficient AI models.

Labeling Experience

French Text & Audio Data Labeling for LLM Training

LabelboxTextEntity Ner ClassificationClassification

Participated in multiple data annotation projects for LLM development, focused on French language tasks. Tasks included: Annotating user intents, entities, and emotional tones in French conversations. Evaluating AI-generated responses for coherence, fluency, and alignment with prompts. Writing high-quality prompt-response pairs in French for supervised fine-tuning (SFT). Transcribing French audio data and applying phonemic labels for speech recognition training. QA reviewing other contributors’ work to ensure consistency and high labeling accuracy. The projects ranged from small batch datasets (under 1,000 items) to large-scale corpora exceeding 50,000 samples. Maintained over 95% quality rating across all assigned tasks and adapted quickly to evolving guidelines and tool updates.

Participated in multiple data annotation projects for LLM development, focused on French language tasks. Tasks included: Annotating user intents, entities, and emotional tones in French conversations. Evaluating AI-generated responses for coherence, fluency, and alignment with prompts. Writing high-quality prompt-response pairs in French for supervised fine-tuning (SFT). Transcribing French audio data and applying phonemic labels for speech recognition training. QA reviewing other contributors’ work to ensure consistency and high labeling accuracy. The projects ranged from small batch datasets (under 1,000 items) to large-scale corpora exceeding 50,000 samples. Maintained over 95% quality rating across all assigned tasks and adapted quickly to evolving guidelines and tool updates.

2023

Education

Work History