For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

Invite to Job
Sophia Saleh

Sophia Saleh

Expert in AI computer vision data labeling

UNITED_KINGDOM flag
Liverpool, United Kingdom
$21.00/hrIntermediateLabelboxSuperannotateOther

Key Skills

Software

LabelboxLabelbox
SuperAnnotateSuperAnnotate
Other

Top Subject Matter

Health Care- Medical records & Patient data
Finance- Risk Analysis and Fraud Detection
E-commerce - Product Categorisation & Customer Support

Top Data Types

AudioAudio
TextText
VideoVideo

Top Label Types

Bounding Box
Classification
Cuboid
Segmentation
Text Generation

Freelancer Overview

Over the course of my roles in data science, research, and quality assurance, I’ve built strong experience working with detailed information and checking it for accuracy, consistency and clarity. In my current position, I regularly review complex outputs, spot errors, and rewrite information so it can stand on its own without extra context. A lot of my work relies on careful reading, precise written communication, and the ability to break down information that isn’t always straightforward. These skills translate directly to tasks such as verifying labels, correcting answers, and producing clean training data for AI systems. My earlier work in programming, academic research, and administrative quality checking has reinforced this attention to detail. I’ve handled technical documentation, extracted information from a variety of sources, and worked independently on tasks that require focus and accuracy. Across all of these roles, I’ve learned how important it is to take a meticulous, thoughtful approach to reviewing information, something that has consistently helped me deliver reliable, high‑quality work.

IntermediateSomaliFrenchEnglish

Labeling Experience

Text Data Labelling & Annotation Quality Review (Academic + project work)

TextQuestion Answering
During my MSc in Data Science and Artificial Intelligence, I worked with multiple text‑based datasets that required careful reading, interpretation, and annotation. In an AI‑bias research project, I reviewed text classification outputs, checked whether model predictions aligned with the intended meaning, and analysed inconsistencies caused by imbalanced training data. This involved manually verifying labels, rewriting ambiguous text, and documenting corrections clearly for both technical and non‑technical audiences. Throughout my academic work, I regularly cleaned and prepared text data for analysis, corrected mislabeled entries, reviewed model‑generated text, and summarised findings accurately. I frequently produced written explanations, project documentation, and evaluation reports, which strengthened my ability to generate clear, concise, standalone text outputs. These tasks required strong attention to detail, precise written English, and the ability to judge whether text should be fixed, discarded, or relabelled skills directly relevant to AI training and text‑labelling workflows.

During my MSc in Data Science and Artificial Intelligence, I worked with multiple text‑based datasets that required careful reading, interpretation, and annotation. In an AI‑bias research project, I reviewed text classification outputs, checked whether model predictions aligned with the intended meaning, and analysed inconsistencies caused by imbalanced training data. This involved manually verifying labels, rewriting ambiguous text, and documenting corrections clearly for both technical and non‑technical audiences. Throughout my academic work, I regularly cleaned and prepared text data for analysis, corrected mislabeled entries, reviewed model‑generated text, and summarised findings accurately. I frequently produced written explanations, project documentation, and evaluation reports, which strengthened my ability to generate clear, concise, standalone text outputs. These tasks required strong attention to detail, precise written English, and the ability to judge whether text should be fixed, discarded, or relabelled skills directly relevant to AI training and text‑labelling workflows.

2022 - 2023

Image data labelling and AI training data preparation (MSc Project)

ImageSegmentation
During my MSc in Data Science and Artificial Intelligence, I worked extensively with image‑based datasets that required manual review, cleaning, and labelling before model training. This included preparing and validating MRI brain‑scan data for a U‑Net segmentation model, correcting inaccurate or unclear labels, checking consistency across samples, and ensuring the dataset aligned with the model’s learning objectives. I also worked on image‑classification projects using CNNs and MLPs, where I reviewed images, adjusted class labels, handled augmentation artefacts, and documented any corrections clearly. Across these projects, I became confident in identifying mislabelled data, assessing annotation quality, and producing reliable training data. I applied evaluation metrics such as precision, recall, and F1 score to judge the quality of labels and model performance. This work strengthened my ability to work carefully and independently, maintain accuracy over repetitive tasks, and ensure data is clean, consistent, and ready for AI training.

During my MSc in Data Science and Artificial Intelligence, I worked extensively with image‑based datasets that required manual review, cleaning, and labelling before model training. This included preparing and validating MRI brain‑scan data for a U‑Net segmentation model, correcting inaccurate or unclear labels, checking consistency across samples, and ensuring the dataset aligned with the model’s learning objectives. I also worked on image‑classification projects using CNNs and MLPs, where I reviewed images, adjusted class labels, handled augmentation artefacts, and documented any corrections clearly. Across these projects, I became confident in identifying mislabelled data, assessing annotation quality, and producing reliable training data. I applied evaluation metrics such as precision, recall, and F1 score to judge the quality of labels and model performance. This work strengthened my ability to work carefully and independently, maintain accuracy over repetitive tasks, and ensure data is clean, consistent, and ready for AI training.

2022 - 2023

Education

U

University of Liverpool

Masters, Data Science and Artifical Intelligence

Masters
2022 - 2023

Work History

H

HMRC

Data Scientist

Liverpool
2025 - Present
H

HMRC

Administrator Officer

Liverpool
2023 - 2025