Projects
With Kayla, Aniket, & Erika - for the DESINST 240 class @ Stanford
Designing ML for the Vaginal Microbiome
We interviewed people with vaginas, across the world, to identify problems they face with sexual health. We designed a platform with an ML-driven UI to cater to varying needs.
Class project for Stanford EDUC 281 (Technology for Learners)
Duolingo Product Analysis
Looking at Duolingo from the perspective of a user and an educator to analyze the learning theories behind the app, to evaluate its pros and cons, and to make recommendations to improve the learning experience.
With Mei, Annie, and Ian-Bott
EdTech Landscape
This study
examines central themes of focus in the edtech industry over the past 40 years and visualizes the evolution of topical distributions over time. In combination, this and future projects aim to define a new taxonomy for edtech and identify areas of disconnect and potential, generating insights for researchers and practitioners to position their work for more significant impact.
With Shubham Rateria, Adri Rajaraman
QuickTri- Non Contact Vital Estimation
The solution aims to address the challenges of overcrowded hospitals during the COVID 19' pandemic by providing vital signs such as heart rate, respiratory rate, and oxygen saturation using the webcam.
@RoboManipal
Autonomous Tetris Solving Robot
Built for the World Robotics Olympiad ( ARC) 18' problem statement: The team worked on the locomotion control of an Omni-wheeled robot, arm alignment and gripping as well as on dedicated mechanisms for other tasks.
With Jithin Sunny
Pedi-Pod (AR-based dosage estimation tool )
It is an AR-based, offline solution for emergency physicians to quickly administer the right dose of medication to pediatric patients by estimating height and weight through a mobile device.
With Ram Rathi, MIT Manipal
DNA Compression with LSTMs
The growing volume of DNA data being generated and stored in databases calls for efficient data compression and storage methods. In this work, we propose a new hybrid algorithm using LSTMs to compress DNA sequences.
Ongoing project with Dr Matias Valdenegro
Modeling Uncertainty in Bayesian NNs
Ongoing virtual summer internship in exploring various algorithms to improve the computational efficiency of Bayesian Neural Networks, particularly w.r.t. Deep Ensembles
Biped by Sanat Mharolkar ( ex-Team Leader, RoboManipal)
Learning Robot Control and Dynamics
Trying to learn MPC, ZMP and other control techniques to take Sanat Mharolkar's project forward and implement closed-loop control, perhaps even RL control algorithms on the 10 DOF biped.