Hafiz Hammad Sharafat Satti Email & Phone Number

Islāmābād, Pakistan

Designed and implemented machine learning models tailored to industrial problem-solving, leveraging data analyticsand predictive modeling to address operational challenges.Hafiz Hammad Sharafat SattNationality: Pakistani Date of birth: 16/08/2000 Gender: Male Phone number: (+92) 3058594110 Email address: hafizhammadsatti@gmail.c Home: House #4 Street #4, Shakrial kurri Road, 46300 Rawalpindi (Pakistan) 1 / 3Developed algorithms that optimized key industrial processes, enhancing productivity, reducing operational costs,and increasing accuracy in decision-making.Collaborated with industry experts to understand specific requirements, adapt models to real-world applicationand provide actionable insights

Islāmābād, Pakistan

Spearheaded a project to develop an AI-based system to assist in heart surgeries, with a focus on analyzing heartregions, estimating fat levels, and correlating these with clinical data.Employed the YOLOv8 model for segmenting and analyzing heart regions in surgical videos, extracting criticalfeatures such as heart area and color to quantify fat presence.Correlated extracted data with patient clinical records, diet, and physical activity using Pearson correlation, aiming toprovide surgical risk assessments and health insights.Built a comprehensive Django-based dashboard that provided an interface for surgeons, enabling them to managepatient profiles, view real-time analysis, generate medical reports, and prescribe medications directly within tplatform.

Islāmābād, Pakistan

Led the development of a real-time computer vision system for Medical Images.Utilizing YOLOv8 to detect and frame images in live video feeds during surgeries.Enhanced model accuracy and performance through advanced data augmentation techniques, resulting in an 85%accuracy on the test set.

Islāmābād, Pakistan

Conducted research on satellite image robustness by strategically adding noise to images to examine the impact onobject detection algorithms, including YOLO, Faster R-CNN, and Mask R-CNN.Analyzed and evaluated various noise types (Gaussian, Salt-and-Pepper, Poisson) to determine their effects detection accuracy.Generated insights into the vulnerabilities of object detection algorithms, contributing to advancements in algorithmrobustness in adverse conditions.Outcomes included identifying specific noise levels that degrade detection performance, laying groundwork ffuture enhancements in satellite image processing