Virtual Audience simulator
Used gesture and voice recognition to build a VR application that can simulate an interactive classroom-like environment. The aim was to help users experience the audience and their response depending on their speech and actions.
The VR experience allows the device to track the user’s position, body movement, speech, expression and gestures in a intelligent responsive environment. One of the most common use case of the application is to rehearse for a presentation in order to overcome stage fright and emulate audience responses.
The application was built and run on an Android device with the help of Unity 3D game engine and Google Cardboard. Based on the data collected, it will demonstrate a virtual audience emulating a response as given by an actual real life audience. The setup may also include some sensors connected to other body parts of the user to detect their complete motion, which can’t be determined by the VR headset itself.
Aasara - Handheld emergency TRACKER
In a team of 4 students from Electronics and Computer Science background, we built a product whose aim was to serve as a safety device for women in rural and underdeveloped areas. Its features included Remote tracking and communication of small network of interconnected devices in an isolated area that can locate and identify themselves when triggered. These devices could be used in dire situations to aid an emergency rescue.
My role was to design the product to bridge user requirements and device capabilities. Apart from writing code for the geo-location of the Arduino nodes, I performed user testing and demand modeling of the device.
The project was developed under the direction of Purdue University through its EPICS (Engineering Projects in Community Service) program where it was shortlisted in top 20 entries from over 100 submissions.
Linksmith.com - online alumni directory
Linksmith is an internal web application for the current Smith school students to gain industry/job role alumni information with the click of a button. It saves the time of googling, and it also provides an option to download the entire information in a user friendly template for later consumption. It was developed using Python’s Flask Web-framework, Microsoft’s SQL server, and Bootstrap CSS and HTML.