• Developed FLEET, a framework-agnostic federated learning testbed integrated with high-fidelity network emulation.
• Enabled scalable, configurable experiments by combining Flower for FL orchestration with Containernet for realistic network environments.
• Supported diverse FL setups including various frameworks, datasets, aggregation strategies, and heterogeneous data distributions.
• Simulated realistic network conditions (bandwidth limits, latency, packet loss, background traffic) and advanced dataset partitioning (IID, shard-based, Dirichlet) to analyze their impact on FL convergence.
• Built comprehensive evaluation tools, including background traffic generation, holistic performance metrics, and reproducible configuration/logging for detailed system and network analysis.

• Focus on optimizing network communication for Federated Learning using Software-Defined Networking (SDN).
• Integrate the Flower framework for Federated Learning with SDN tools like ONOS, Mininet, and PyTorch to enhance the efficiency and scalability of distributed machine learning models.
• Utilize the SDN controller (ONOS) to manage dynamic routing, minimizing communication latency and improving performance across multiple edge devices.
• Test in simulated environments using Mininet and optimize through real-time

As the sole backend developer with 600+ commits since 2022, I designed and built the core infrastructure of Wird, a student Ramadan activity tracking platform supporting 1,000+ daily users. The application enables admins to create customized contests, track student progress, and export detailed reports, while students log activities and compete in a gamified environment.

• Backend Development: Built scalable Django REST Framework (DRF) APIs powering both a React-based admin panel and a Flutter mobile app for students.
• Performance Optimization: Implemented Redis caching and PostgreSQL optimizations to enhance speed, scalability, and reliability.
• Impact: Delivered a seamless system that supports contest management, student tracking, and real-time progress monitoring, driving high user engagement.

• Designed and developed RENET, an adaptive routing algorithm to handle bandwidth fluctuations in edge networks, with a focus on remote areas and challenging weather conditions.
• Implemented a delay-based measurement technique for rapid detection of wireless link fluctuations and dynamic flow path adjustment to maximize network utilization.
• Achieved 30% higher QoS satisfaction rates compared to existing solutions and improved adaptive video streaming quality by up to 54%.
• Validated performance in both real-world deployments and controlled emulated network environments using diverse testing tools.

Developed a personal portfolio website using Wagtail CMS, designed to showcase research, publications, and projects in a clean, professional, and easily maintainable format.

• CMS & Backend Development: Leveraged Wagtail on Django to create a flexible content management system with custom models and reusable components.
• Design & Usability: Built a responsive, modern interface highlighting academic and professional work, optimized for clarity and accessibility.
• Scalability & Maintenance: Structured the site for easy content updates, ensuring long-term maintainability and extensibility for future features.