Islam Choriem Cheiref

Key Skills

Software

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Video

Top Label Types

Audio Recording

Transcription

Freelancer Overview

Iam a trilingual generalist with hands-on experience in AI training and data labelling, I have worked on projects to enhances LLMs models accuracy and efficiency. I have been involved in curating and validating multiple high quality datasets, ensuring consistency and reliability across different languages. I have developed an ability to follow complex guidelines and a keen understanding of how well labeled data impacts AI performance In addition to my languages skills, that consist of natively speaking both arabic and French, and being an advanced English speaker with a 7.5 band IELTS certificate (C1), I posses fondamental knowledge in the computer science field as part of my Bachelor degree, allowing me to collaborate effectively with technical terms and understand the data pipelines behind AI -systems. This combination enables me to be a valuable contributor to AI training initiatives involving diverse datasets and multilingual environments.

Labeling Experience

Xylophone Grassland

Scale AIVideoTranscription

A multimodal data collection project where short improvised speech were recorded in various environments and then transcribed, emphasizing on speech capture, clean transcription and accuracy to describe the acoustic conditions; in goal to support robust speech recognition and audio language model training in different languages

A multimodal data collection project where short improvised speech were recorded in various environments and then transcribed, emphasizing on speech capture, clean transcription and accuracy to describe the acoustic conditions; in goal to support robust speech recognition and audio language model training in different languages

2025 - 2025

Xylophone Conversation

Scale AIAudioAudio Recording

A large scale multilingual project consisiting of recording conversations and human dialogues in high quality on various subjects to train LLMs to replicate human-like tonality and naturalness in its vocal replies

A large scale multilingual project consisiting of recording conversations and human dialogues in high quality on various subjects to train LLMs to replicate human-like tonality and naturalness in its vocal replies

2025 - 2025

Education

Work History