AI-powered glasses (using computer vision) for the Visually Impaired to aid in Question/Answer, Navigation, Reading, and Recognising text/currency (patent pending), all while costing less than $20. Received large grant from IBO, IRIS National Fair Finalist. 

"Drawing in the air!" I worked with computer vision techniques to create hand-tracking AR/VR drawing software for accessible immersive experiences, with future applications in engineering and shared XR experiences.

Using statistical analysis to establish novel rank-size distributions in road networks, suggesting Data Augmentation to improve AI model function in high-uncertainty situations. This paper is currently under review.

I developed a custom-made general-purpose AI model with characteristics of multi-layer regression and KNN models. This allows for the weightage of input factors and, therefore, dimensionality reduction.

I developed gloves that mimic realistic haptic feedback but with AR/VR content, using image tracking methods and wireless microcontroller communication. Applications in the future include remote surgery and training.

For users with disabilities (and occupied hands), I created a functional webviewer to use facial features to scroll webpages, as an iOS application. Other actions like specific blinks could interact with buttons.

Salvaging broken parts and a damaged flight controlled, I created a lightweight drone to learn more about the integration of hardware and software.

In-depth computational research into algorithmic simplification and base-2 representations of values in Collatz orbits (and some stochastic patterns). Submitted to American Mathematical Monthly and winner of Grand Award at IRIS 2024-25.

Addressing the issue of high urban fire casualty rates and improper access mechanisms in many cities, I developed and prototyped a low-cost drone which can autonomously detect and mitigate fire hazards.

I have developed 6 iOS applications (on the app store) to tackle various problems ranging from inefficient unit conversion to AR annotations and offline map functionality. More information in the 'Apps' section.

Custom-made 3D printer implementing novel features, including pre-extrusion mechanisms (to minimise filament breakage). This allows cheaper 3D printing and 3D printing of more fragile (even biocompatible) materials.

I designed a small and 3D-printable electron accelerator in the form of a small Cathode Ray Tube constructed from a damaged plasma globe. This can be extended to an electrode-based linear accelerator. 

Collaborating with 4 others, I worked on addressing the problem of improper solid waste management in Bangalore's slums through an accessible lightweight application, providing monetary incentives to waste disposal.

I experimented with creating a complete application using LLMs to offer personalised mental health feedback and online sentiment analysis for various social media platforms (as an iOS application).