I interned at Open Water Accelerator this summer. Meeting emerging founders and collaborating woth fellow interns to help portfolio company Dcyphr develop technology beyond the one used by it's competitors was once of the best Experiences of my life.

I programmed a Bidirectional Recurrent Neural Network for abstractive text summarization using Facebook’s BART model to simplify medical papers for researchers who spend up to 11 hrs. reading journals. By using attention models to generate summaries of scientific papers, I improved accessibility of scientific literature for the common man. I also played around with applying transfer learning to PEGASUS to generate simplified summaries of research papers

Research is fun! I actually took two research projects this summer. I worked on the Xenophobic Meter Project for the Cornell Law School, where I developed an NLP application to detect racism and xenophobia in tweets. We initially used the Naive Bayes Algorithm to perform sentiment analysis, but then we decided to switch to using Recurrent Neural Networks and ended up using Google's BERT to give a toxicity score to tweets. I have been fortunate enough to work with Beth Lyon,the dean of the Cornell Law School.

My second undergrad research was for Cornell College of Engineering. Unsafe drinking water is a major problem in developing countries.We developed a scalable solution using Python, Featurescript and OOP principles by automating a design engine that can regenerate designs of industrial infrastructure after the user provides input to it. We developed geometry to model sedimentors, flocculators and filters and write code to customise the configuration of the industrial infrastructure given expert inputs by the user. Our minimum Viable Product is a 3-Level build feature that regenerates up to 90% of the Onshape drawing. I worked under Professor Monroe Weber Shirk and he is amazing!