I ♥ technology, and am passionate about developing the tech of tomorrow.
I am a research scientist at Google, focusing on health sensing hardware. My interests lie at the intersection of sensing research and systems engineering.
Previously, I have worked at a silicon valley startup (Wisk Aero) evaluating sensing for autonomous eVTOL flights and also at an early phase startup (DeepSig) creating AI native communications stack.
As a master's student in Computer Engineering at Virginia Tech, I dabbled in
neuroscience research with Prof Sujith Vijayan on
Brain Computer Interface.
I also like making fun stuff with a raspberry-pi,
playing the piano, going on long hikes and contributing to open source!
Research Scientist
Health sensing research, focusing on optical signal processing and algorithms for heart rate and PPG signal sesning..
Software Engineer - Embedded Systems
Sensors, flight systems and writing embedded code that flies!.
undefined
Flying taxis. Yeah! I am equally amazed! Check them out here.
My work involves sensor integration and testing (and also designing tests, performing Data Analysis) to meet aviation requirements. I work on a novel aviation sensor solution and making sure it works the way we expect it to! My mentor usually throws a new piece of hardware on me, for testing/bringup and interfacing with aviation hardware (basically the code I write is flying high!). .
Machine Learning Intern
I am working on interfacing the software suite of the company with different OpenSource tools for improving the signal processing and the machine learning workflow.
I am also working on the developement of the OmniPhy product, which is an optimised wireless channel encoder-decoder (PS. Think 6G comms)..
Graduate Student Researcher
I am working on developing a deep neural network for modelling the brain activity. Currently we are experimenting with different architechtures such as LSTM+RNN, Spectograms as Images and also traditional ConvNets.
I am also working on interfacing a BMI(Brain Machine Interface) using EEG and ECoG data to study the effects of sleep. I am also building different classifiers for complex BMI tasks. (Mind control. Works. Mostly.).
Sofware Engineer
I was initially working on a navigation problem for autonomous driving in various traffic scenarios.
Later, I joined the Daimler Trucks Powertrain Diagnostics Team where I was working on ECU measurement, calibration for the Euro IV,VI and the NAFTA market. I also worked on Test Automation for Diagnostics Testing to reduce time spent in Freeze Frame Testing and ISO27145 Protocol conformance Testing.
.Android Developer
I was reponsible for building and integrating data capture libraries for images, videos, QR codes and text inputs for managing health records for pets. The interface designed was aimed to user-friendly and interactive. I also worked on designing the SQLite database to manage and store the records.
.Embedded System Developement Intern
I worked on the challenge of simultaneous localization and mapping for an autonomous agent whose only sensor input was an Inertial Motion Unit. The project enhanced the reliablility and accuracy of a single sensor system and overcame the partial drawbacks of using accelerometers in an Inertial Navigation System.
.IOT product design intern
My team worked on designing a wireless video doorbell service for accessible and portable use. I worked on Python based client-server interaction for a storing and managing the video stream data based on the doorbell input. The accessiblity was increased by developement of a Web App and an Android App.
.Unique interaction media, helping you stay in touch with your loved ones. Ping me if you're interested in working on this!
Inspired by Simone Giertz Yearly Calender, I created a DIY habit tracker using a raspberry pi. It encourages and enables me to keep a track of my weekly progress of any two habits.Blog post for the same here
A Minimalisitic Navigation system designed and implemented for bicyclists. It provides direction and distance to the next turn using a intuitive LED interface. Modular. Cheap. Simple. Looking for investors....
AR Simulation was induce the visual and auditory hallucinations caused by Schziophrenia. The experiment was aimed at educating the general masses about the disease to reduce the stigma associated with the same.
Find a link to the experiment here
Play a game of pong using the beta rhythm from the brain activity using various EEG headsets. The Foce is strong within this one!
Using histogram of oriented gradients feature along with a SVM classifier, we trained a model for gesture recognition for Finger Spelling. A predicited word output was also given using the letters predicted from this model to increase letter coverage and accuracy for real time video camera input. A Deep Learning model was also trained to perform the same task to compare the two models.
Generate a thug lyf image from your existing images using MATLAB. Uses the MATLAB face and eye detector to add the thug lyf shades on the face because why not!
We were given admin logs(~20GB) from various machines from a company. We had to find evidence of malware and keylogging attacks. I also worked on predicting employee resignations and model suspicious activity using various analytical models.
Implemented a k-d tree classifer to cross match galaxies from the SuperCOSMOS dataset to the ATG20 BSS catalog. Also, built CART trees from the SDSS dataset and used ensemble learning to classify galaxies according to the Hubble Classification Chart.
Image Credit : Sloan Digital Sky Survey>Some useful/fun web spiders and scrappers that contains scripts used to index, learn from and explore various web sites and content.
Trello Integration - Made an on-desk gadget to keep of your to-dos!Placed 4th among 247 teams.(National Event)
Implemented PID controller on FRDM KL25Z µC for steering a model car across turns, u-turns, bends and bumps
Placed 3rd among 50+ teams(National Event)
Design of co-ordinated Autonomous and a semi-autonomous robot having 6 degrees of freedom. I worked on developing it's navigation system and automating it to perform various tasks. I also worked on driver level programming for I2C and SPI bus connecting a system of 7 uC.
Tips and Tricks that will help you stand out when applying for a job..
My experience of working at a Bay area startup making flying taxis!.
HabitPi - Daily Habit Tracker I made using Raspberry Pi. The blog describes the thought behind the idea and its inspiration..
Real Time Operating Systems
Career Developement - Professional
Experimental Robotics (Audit)
Wearable and Ubiquitous Computing
Virtual Enviornment
Advanced Parallel Computing
Electronic Design Automation
Computer Vision
Data Analytics
Network Architechture and Protocols
Digital Image Processing
Automotive Engineering
Speech and Audio Processing
Control Systems