PARIJA MALGAONKAR

Comparison of Kalman Filtering Techniques in Inertial Navigation Systems:

analyzed Bayesian state estimation methods to improve sensor fusion and state accuracy in UAV-based Strapdown Inertial Navigation Systems (INS). The work involved modeling INS dynamics, transfer alignment, and GPS-INS integration, and implementing four filters: Simple Kalman Filter (KF), Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), and Particle Filter (PF). I developed nonlinear state-space models, derived Jacobians for EKF, applied the unscented transform for UKF, and implemented importance sampling and resampling for the Particle Filter while evaluating covariance propagation and predictor–corrector updates. Using MATLAB and Python, I performed Monte Carlo simulations and comparative performance analysis to assess estimation accuracy, robustness to nonlinearity, and computational complexity.

Mind Mapping Web App: 

ReactJS, NodeJS, MongoDB, FlowpointJS, Android Studio, Java, GoogleAuth. 

Built a real-time mind-mapping tool with ReactJS and FlowpointJS used by 1K+ users for visual idea organization and drag-and-drop node mapping. Developed a Node.js + MongoDB backend supporting autosave, data persistence, and cross-device sharing. Integrated GoogleAuth across web and Android (Java, Android Studio) for secure OAuth2 login, reducing failures by 20%. 

E-Commerce Web App (Online-Offline Retail): 

ReactJS, NodeJS, MongoDB, AirtableJS, Flutter, Java, Firebase Authentication.

Built a scalable cross-platform e-commerce app with ReactJS, Flutter, and Node.js, supporting 10K+ products and 5K+ daily users. Managed dynamic inventory and seller data using MongoDB and AirtableJS, streamlining sync across online and offline partners. Enabled secure real-time appointment booking and store recommendations via Firebase Auth, geolocation, and behavior-driven logic.