Projects

2024
Outlier.ai
Data Annotation
As an expert contributor on Outlier.ai, I have had the opportunity to work on a wide range of data annotation tasks. Each project presents its own set of challenges and learning opportunities, keeping the work engaging and fulfilling. The platform’s emphasis on RLHF and SFT methodologies ensures that every contribution directly impacts the performance and quality of the AI systems we help train.
2023
Birmingham City University
Medical Image Segmentation
As my dissertation for my Master's degree at Birmingham City University, I implemented U-Net and Duck-Net deep learning architectures for precise medical image segmentation using Python and compared both convolutional neural networks CNNs on the CVC-ClinicDB dataset available through Kaggle. Yielded a Dice Coefficient of 0.91 with slight modifications to the original Duck-Net architecture
2021
Style Textile (Pvt) Ltd.
Heat Recovery Unit Erection and Automation
At Style Textile, I led the automation of a heat recovery unit to enhance energy efficiency in the dyeing process. Using advanced PLC and HMI systems, I designed and implemented an automated control mechanism to recover heat from waste warm water and use it to preheat incoming fresh water. This innovation reduced the facility’s reliance on conventional energy sources, resulting in annual savings of over £1 million in coal costs and a 20% reduction in carbon emissions. Collaborating with German engineers during commissioning, I ensured adherence to international quality standards, showcasing my ability to deliver sustainable engineering solutions.
2021
Style Textile (Pvt) Ltd.
Dyeing Plant Automation
Automated the plant's operational processes using Orgatex and Lawer systems for automated dye and chemical dispensing. Implemented real-time monitoring systems for predictive maintenance and operational efficiency. Reduced production downtime and improved operational accuracy through data-driven solutions.
2020
Style Textile (Pvt) Ltd.
Machine Erection and Commissioning
Led the commissioning of several high-tech dyeing machines, including Orgatex and Lawer systems, ensuring smooth integration with existing systems. Coordinated with vendors and internal teams to complete the project 20% ahead of schedule, reducing downtime by 25%. Utilised data analysis techniques to improve machine uptime by 30%, enhancing production capacity and reducing operational costs.
2019
National University of Science and Technology (NUST)
Automated Identification of Abnormal EEG
As my final year project for my Bachelor's degree at NUST, I did research on enhancing the accuracy of the chronoNET algorithm for identifying abnormal EEG patterns using deep learning techniques. This research contributed to advancements in EEG analysis using Deep Neural Networks