http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/36 2026-04-28T12:56:58Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17117 MediSync BD - An Innovative Telemedicine Platform with Real Time Video Consultation Mohiuddin, Muhammad MediSync - An Innovative Telemedicine Platform with Real Time Video Consultation is a complete digital health platform aimed to connect patients and health institutions in real time video consultation and built-in appointment management. The platform will overcome traditional healthcare access disadvantages, including geographical inaccessibility and long queues as well as administrative wastages. MediSync provides its patients with a friendly web and mobile platform to schedule virtual appointments, talk to doctors through live video, read their electronic health record, and have follow-up care - all in the comfort of their homes. To healthcare providers, the system provides them with a centralized platform to manage schedules, make online consultation, and access patient data in a secure and efficient manner in service delivery. Its major characteristics are real-time video conferencing, creation of eprescription, appointments bookings, and patient data storage in the cloud, which adheres to the healthcare data standards of data protection. MediSync intends to increase patient accessibility, healthcare efficiency, and quality care delivery at all times and in all locations by incorporating telemedicine with modern web technologies. The main mission of MediSync is to redefine healthcare experience through developing a more connected, accessible, and patient-centered health care system - a gap bridging between patients and healthcare providers in the age of digitalism Project Report 2025-09-24T00:00:00Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17116 Deepfake Image Detection Using Deep Learning Models Das, Sajib Fake images are becoming a threat to our society. Recognizing fake images with the naked eye is a very difficult task. And in many cases it is impossible to recognize even after much effort. Due to which it is becoming easier to spread false news among people and misguide them with less effort. This problem will become even bigger in the future and will pose a threat to human safety. In this research paper I performed deep learning models one by one and from those I took the best performing models. With their help, I created a hybrid model using the best performing models. My goal is to create a hybrid model that will give high accuracy and high recall rate. And will also perform well on any custom dataset. To this end, a custom hybrid model was created and implemented leveraging innovative factorized residual blocks to achieve efficient feature extraction with reduced parameter counts. Firstly I performed with single models MobileNetV2, InceptionV3, EffieientNetV2B0, ResNet50, VGG16. From these models I took 2 models with best accuracy, precision, recall and f1 score and combined them. Then I fine-tuned for binary classification of real versus AI-generated images. This hybrid model trained and validated on the CIFAKE dataset, which contains labeled samples of authentic and synthetic images. I made a custom dataset also to check how my hybrid model on unseen dataset and it performed pretty well. Training leveraged GPU acceleration within TensorFlow/Keras frameworks to optimize computational performance. The model performance results are presented in the form of accuracy, precision, recall, f1 score, performance metrics and confusion metrics generated by the model. And I saw that Mobilentiv2 and ResNet50 are performing the same results on the CIFAKE dataset. And the rest of the models are performing almost close. So I created a hybrid model using MobilityTV2 and ResNet50 as the best performing single model. The hybrid model performs well. Robustness check is done by using a custom dataset and checking the unseen dataset. For the CIFAKE dataset my hybrid models accuracy was 0.98 and recall rate also increased to 0.98. And for the custom dataset my accuracy was 0.8875 and recall rate was 0.8875. This study underscores the feasibility of using hybrid models for fake image detection, providing a practical pathway toward scalable, real-time systems capable of mitigating the spread of AI-generated misinformation. Thesis Report 2025-08-11T00:00:00Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17114 AI-Driven DevOps: Enhancing Automation, Efficiency, and Reliability in Software Development Ratul, Radwanul Habib We are seeing that software development is changing very fast, and it demands only faster and more secure delivery methods. Traditional DevOps practices automate workflows effectively, but they have limitations in predictive analytics and anomaly detection itself. Further, these practices cannot make intelligent decisions automatically. This thesis studies how to add AI into DevOps work processes, focusing on CI/CD pipelines, as per the need to improve automation and efficiency. The research examines AI integration, which means better workflow management and system strength. Usually, the research visualizes a complete method for creating and testing AIpowered CI/CD systems that helps to cover the some steps like data collection, model training, and monitoring. The methodology includes all essential processes from preprocessing to evaluation and explainability. We use real CI/CD log data to create AI-powered workflows with machine learning models. These models definitely help with automatic testing, better deployments, predicting failures, and finding security problems. The dataset is further divided into training, validation, and testing subsets for proper model evaluation. The model itself achieved 100% accuracy, precision, and recall in simulated experiments. As per comparative analysis, AI-driven DevOps shows better results than traditional DevOps regarding build success rates, deployment speed, recovery time, and security detection. When we do AI integration is used in DevOps processes, it helps to reveal significant improvements. Case studies in the real world show that AI methods work well in large software systems. Moreover, these approaches help reduce system failures, fix problems faster, and make security better. This research surely helps connect traditional automation with smart decision-making in DevOps. Moreover, the findings bridge an important gap between these two approaches. Basically, when organizations add AI models to their CI/CD pipelines, they get the same result - software delivery that can fix itself, predict problems, and adapt automatically. The study surely concludes that AI-driven DevOps is not just an improvement of existing practices. Moreover, it represents a complete transformation toward fully independent software engineering systems. Thesis Report 2025-09-11T00:00:00Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17112 Essenza - A Clothing eCommerce Store Fahim, Omar Faruqe This project, Essenza – A Clothing eCommerce Store, is an online shopping platform designed to sell clothing and related products with a focus on user-friendly design and automated bundle offers. The system provides essential e-commerce features such as user registration, product browsing, cart management, secure checkout, multiple payment methods, and order tracking. It also includes an admin panel for product and sales management, as well as real-time discount offers to encourage bulk purchases. The platform has been developed using React.js for the frontend, Node.js for the backend, and MongoDB for the database. Agile methodology and the Software Release Life Cycle (SRLC) have been followed for development and deployment. Comprehensive testing strategies, including functional, performance, and security testing, ensured system reliability, scalability, and compliance with privacy standards. Essenza successfully supports up to 2,000 concurrent users with fast response times and achieved positive results in usability and performance. Future enhancements include AI-based personalized product recommendations, augmented reality (AR) try-on features, and multilingual support to reach a wider audience Thesis Report 2025-08-11T00:00:00Z