Ines Abdelaziz

Key Skills

Software

AWS SageMaker

CrowdSource

CVAT

Labelbox

Label Studio

Prodigy

SuperAnnotate

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Classification

Computer Programming Coding

Data Collection

Prompt Response Writing SFT

Text Generation

Freelancer Overview

I have extensive experience in AI training data and data labeling, with a focus on Natural Language Processing (NLP), text classification, and automation of labeling workflows. I’m skilled in creating and labeling datasets for tasks like sentiment analysis, named entity recognition (NER), and text generation. My proficiency in programming, particularly Python, allows me to automate repetitive tasks, ensuring data quality and efficiency in large-scale labeling projects. I’ve worked with tools like Label Studio and Labelbox, optimizing data pipelines for various machine learning tasks. In addition to NLP, I have hands-on experience in e-commerce and customer analytics, where I’ve contributed to building recommendation systems and sentiment analysis models. I also have expertise in creating prompt/response pairs for training chatbots and dialogue systems, which are essential for enhancing customer interaction experiences. My ability to combine coding and data labeling enables me to streamline the AI training process, reducing errors and increasing productivity. I specialize in making large datasets ready for use in applications like automated content moderation, predictive analytics, and AI-powered customer service tools.

Labeling Experience

Image Annotation for Product Recognition in E-commerce

SuperannotateImageBounding BoxSegmentation

In this project, I worked with an e-commerce platform to annotate over 100,000 product images for use in a computer vision model designed to improve product search and recommendation algorithms. The project involved labeling images for object detection and classification (categorizing products into predefined classes). I also worked on semantic segmentation to label specific regions of product images, such as logos or product labels, which was critical for a brand recognition system. I used Labelbox for overall image classification and SuperAnnotate for more complex tasks like object detection and segmentation. To ensure the accuracy of the annotations, I implemented a tiered review system where team members cross-validated labels, with random checks on high-priority items to maintain quality. The final dataset was used to train a deep learning model that enhanced the platform's ability to recommend and search for products based on visual attributes.

In this project, I worked with an e-commerce platform to annotate over 100,000 product images for use in a computer vision model designed to improve product search and recommendation algorithms. The project involved labeling images for object detection and classification (categorizing products into predefined classes). I also worked on semantic segmentation to label specific regions of product images, such as logos or product labels, which was critical for a brand recognition system. I used Labelbox for overall image classification and SuperAnnotate for more complex tasks like object detection and segmentation. To ensure the accuracy of the annotations, I implemented a tiered review system where team members cross-validated labels, with random checks on high-priority items to maintain quality. The final dataset was used to train a deep learning model that enhanced the platform's ability to recommend and search for products based on visual attributes.

2024 - 2024

Text Classification & Sentiment Analysis for Customer Feedback

LabelboxTextEntity Ner ClassificationClassification

For this project, I worked with a large dataset of customer feedback (reviews, surveys, and social media comments) from an e-commerce platform. The goal was to label the text data for multiple purposes: sentiment analysis, product categorization, and entity recognition (e.g., identifying product names, brands, customer pain points). The scope included over 50,000 customer reviews that required tagging for positive, negative, and neutral sentiments, as well as classifying text based on specific product categories (e.g., electronics, apparel). Additionally, I incorporated NER to extract key entities like product names and features. I used Labelbox for large-scale classification. I developed a custom Python script to automate initial labeling for certain categories, increasing productivity and ensuring consistency across the dataset. Quality control measures included regular team reviews, multiple rounds of validation, and cross-checking for labeling accuracy.

For this project, I worked with a large dataset of customer feedback (reviews, surveys, and social media comments) from an e-commerce platform. The goal was to label the text data for multiple purposes: sentiment analysis, product categorization, and entity recognition (e.g., identifying product names, brands, customer pain points). The scope included over 50,000 customer reviews that required tagging for positive, negative, and neutral sentiments, as well as classifying text based on specific product categories (e.g., electronics, apparel). Additionally, I incorporated NER to extract key entities like product names and features. I used Labelbox for large-scale classification. I developed a custom Python script to automate initial labeling for certain categories, increasing productivity and ensuring consistency across the dataset. Quality control measures included regular team reviews, multiple rounds of validation, and cross-checking for labeling accuracy.

2024 - 2024

Education

Work History