Khaoula Meliane

Key Skills

Software

Remotasks

Scale AI

Top Subject Matter

Response evaluation and correction- Image, text and audio

Text generation in different themes and formats

Ranking and classification

Top Data Types

Audio

Image

Text

Top Task Types

Classification

Evaluation Rating

Prompt Response Writing SFT

Text Generation

Text Summarization

Freelancer Overview

As an AI trainer, I have contributed to various projects that demand high analytical precision, a creative approach, and strong linguistic expertise. Over time, I have developed my skills in refining AI responses to make them more accurate, human-like, and efficient in responding to user queries. My experience encompasses key tasks such as text generation, audio transcription, response evaluation, and correction, which have allowed me to deeply understand AI behavior and improve its performance. These experiences set me apart, equipping me with the adaptability and insight needed to elevate the quality of AI training data.

Labeling Experience

AI trainer

Scale AIAudioQuestion AnsweringTranslation Localization

In this project, I hand several tasks at assessing audio quality and relevance. My responsibilities included evaluating and scoring audio outputs based on a range of criteria, verifying that each audio clip accurately met the user’s request, and comparing multiple AI-generated audio samples to identify the best match. Additionally, I ensured that each audio file adhered to specific requirements, such as clarity, tone, pacing, and overall fidelity to user expectations. This process involved detailed analysis and comparison, enhancing the accuracy and effectiveness of AI-generated audio responses. Also, I worked on recording a transcribed text to generate an audio that meets the local criteria.

In this project, I hand several tasks at assessing audio quality and relevance. My responsibilities included evaluating and scoring audio outputs based on a range of criteria, verifying that each audio clip accurately met the user’s request, and comparing multiple AI-generated audio samples to identify the best match. Additionally, I ensured that each audio file adhered to specific requirements, such as clarity, tone, pacing, and overall fidelity to user expectations. This process involved detailed analysis and comparison, enhancing the accuracy and effectiveness of AI-generated audio responses. Also, I worked on recording a transcribed text to generate an audio that meets the local criteria.

2024

AI trainer

Scale AIImageQuestion AnsweringText Generation

In this project, I was responsible for several tasks focused on image quality and relevance. These included evaluating and scoring the quality of images based on various criteria, verifying that each image accurately aligned with the user's request, and comparing two or more AI-generated images to determine the best answer. Additionally, I ensured that each image met specific requirements, such as style, composition, and clarity, to deliver results that fully adhered to user expectations. This process involved a thorough analysis and comparison to refine AI image generation and fulfill user satisfaction.

In this project, I was responsible for several tasks focused on image quality and relevance. These included evaluating and scoring the quality of images based on various criteria, verifying that each image accurately aligned with the user's request, and comparing two or more AI-generated images to determine the best answer. Additionally, I ensured that each image met specific requirements, such as style, composition, and clarity, to deliver results that fully adhered to user expectations. This process involved a thorough analysis and comparison to refine AI image generation and fulfill user satisfaction.

2024

AI trainer

Scale AITextClassificationText Generation

In this project, I had several tasks, including generating responses, stories, advice, and other content, then comparing them to the responses generated by the LLM to surpass the state-of-the-art and provide the ideal answer to each prompt. Additionally, I localized each response to align with the specific country and cultural context, classified answers into different categories, summarized texts and articles, and transcribed content with an altered tone or context. Finally, I also evaluated and compared two responses based on criteria such as accuracy, localization, and adherence to the given instructions.

In this project, I had several tasks, including generating responses, stories, advice, and other content, then comparing them to the responses generated by the LLM to surpass the state-of-the-art and provide the ideal answer to each prompt. Additionally, I localized each response to align with the specific country and cultural context, classified answers into different categories, summarized texts and articles, and transcribed content with an altered tone or context. Finally, I also evaluated and compared two responses based on criteria such as accuracy, localization, and adherence to the given instructions.

2023

Education

Work History