I am currently a software engineer at Academia.edu.
In summer 2020, I worked on machine learning, where I made a recurrent neural network that predicted whether people will get glaucoma. It had an 81 percent accuracy on a test set where half the people later got glaucoma and the other half did not. We had approximately one hundred thousand people in the data set, and we controlled for age, gender, race, and ethnicity.
In summer 2019, I used machine learning for time-series analysis. Oracle runs a cloud computing service, and the company buys new server parts every quarter. To buy the right number of parts, the company needs to predict how much traffic they will get over the next quarter. I wrote a deep learning model using RNNs and one-dimensional CNNs to forecast server traffer over the next 90 days. In my internship's last week, Oracle used my model to decide how many GPUs to buy.
In summer 2018, I worked on machine learning, where I helped a more senior data scientist develop a deep learning model to classify human activity. Here, I learned the basic tools of machine learning, like Keras and sklearn.
I usually use React with Redux. For example, I used them to help my friend develop Guesstimoji, a web game. The code is available on GitHub. I'm also comfortable writing CSS; for example, I wrote the CSS for this website.
I used Python with Keras for machine learning in my internships. I'm learning Pytorch now, which I hope to use more in the coming years. I do data analysis using the standard data science tools, such as pandas and scipy.
I graduated from Rutgers University with a BS in computer science and mathematics.
In my free time, I like to lift weights, skateboard, bike, swim, and play other sports and activities. I also like to travel; I learned Mandarin during the 2020 quarantine, and I hope to travel abroad to Mandarin-speaking places soon. I also speak enough Spanish and Hindustani to travel.