At Samsung, my work involves designing and developing scalable & robust software solutions in connectivity infrastructure.

Full Stack Application Development- Contributed to the development of ADP’s unified recruitment management product for candidate and recruiter personas.- Leveraged Angular and AWS tech stack for API development & deployment.

Project: Scheduling and Resource allocation (SRA) for OFDMA in IEEE 802.11 ax (NP Hard problem)- Built a generic low latency scheduler, and a resource allocator to optimally allocate limited resources to unbounded number of users. - Developed a monitoring ecosystem to collect, process and analyze key performance indicators for operational review.- Improved KPIs: Reduced latency by 90%, improved runtime by 2.25x, and minimized the max. OFDMA frame transmission time by 91.3%.- Benchmarked with state-of-the art schedulers: Max. Sum Rate, Proportional Fair, Shortest Remaining Processing Time.

Project: Connect the WiFi-based IoT Nodes in Challenged Scenarios facing intermittent connectivity.Developed an Association Free protocol adaptive to varying channel load conditions, improving the goodput by 10x, delivery rate up to 77%, and scalable upto 300 nodes.

Project: Lungs diagnosis using Evolutionary Algorithms and Machine Learning— Architected an automatic lung disease detection tool for early diagnosis of COPD and Pulmonary Fibrosis from CT scan images (lung segmentation, feature extraction and selection, classification).— Modeled and applied bio-inspired algorithms (Crow Search, Grey Wolf) to find optimal subset of features, increasing classification accuracy by 21%. — Deployed the framework for real-time delivery of analysis at university hospital, leading to computational cost reduction by 35%.Published Paper:

Project: Resource allocation and sharing in Flying AdHoc Networks (FANETs)Devised a novel token based fault-tolerant algorithm to resolve classical resource allocation problem in distributed Flying Ad Hoc Networks (a variant of Mutual Exclusion problem and an extension of dining philosopher problem).Published Papers:

— Designed, Implemented and Tested Classplus’ tutor-student-parent app using Model View Presenter (MVP) architecture and Dagger2 Dependency Injection Framework.— Engineered a push-notification engine and report generation framework for 400K+ Daily-Active-Users (DAU) of platform.— Designed A/B experimentation system with inbuilt exposure logging used to run 10+ experiments a day.— Implemented features such as broadcast message, online attendance, announcements, automatic online test creation, test report generation, performance analytics for Classplus - a tutor-student-parent communication enabler platform.— Developed doodling feature used in chat messenger and assignment checking. Worked on features like draw, edit, write on images, redo and undo features, used by more than 200k users.