Positions
ML Software Engineer | October, 2020 - Present
I am currently an ML Software Engineer at Amazon as a part of the Alexa team. I enable Alexa to talk in different voices and languages (Alexa can speak in 16 languages as of 2022!). I have built several voices and language models, as well as productionised various other ML models built for Text-to-Speech (TTS) problems within Alexa.
I also worked at Amazon Web Services (AWS), at Managed Streaming for Apache Kafka (MSK). I built features using various AWS technologies to provide Apache Kafka as a fully managed service on AWS. The service automatically provisions and manages the resources necessary to provide real-time data streaming capabilities. I worked with hundreds of thousands of clusters to scale resources instantly to meet various applications' demand.
Gradate Student | May, 2020 - September, 2020
I worked with Dr. Paul Goldberg and Dr. Alejo Nevado-Holgado, along with Dr. Wilfried Haerty and Dr. Elizabeth Tunbridge for my Master's dissertation.
I explored the structure of the human genome by training neural network architectures to distinguish between DNA sequences containing different types of genomic elements (intron, exons, splice junctions), as well as to predict the exact positions of splice junctions in DNA sequences. I introduced DeepDeCode, an attention-based deep learning model to do this. I received a Distinction in my dissertation.
Research Engineer | August, 2018 - July, 2019
I worked with Dr. Wynne Hsu and Dr. Lee Mong Li at the Institute of Data Science, NUS.
The first project I worked in was on estimating uncertainty of neural networks for medical imaging to detect Diabetic Retinopathy. I also created visualizations to find the areas of the image the neural network uses to make its predictions to build trust the results of a deep learning system.
I also worked on Named Entity Recognition of medical concepts from Electronic Health Record (EHR) data in an unsupervised setting using a hybrid Bidirectional LSTM-CNN model.
Research Intern | May, 2018 - July, 2018
I worked with Dr. Serge Iovleff as a part of the MODAL (Models for Data Analysis and Learning) Research team.
I was one among 13 students selected from 1000+ applicants in India for the highly competitive Charpak Research Internship Program. I worked at the National Institute of Research in Computer Science and Automation (INRIA) on the CloHe collaborative project (funded as part of the CNRS Mastodons 2016 challenge call). The project aimed to develop, study and implement supervised and unsupervised classification methods on satellite and aerial images of France, when the data is of different natures and have missing and/or aberrant data.
Research Intern | May, 2017 - June, 2017
I worked with Dr. Yi-Hsuan Yang in the Music and Audio Computing Laboratory.
I was one among 16 students selected among 5000 international applicants worldwide selected for the International Internship Program under Taiwan International Graduate Program (TIGP-IIP). I worked at Academia Sinica along with our collaborator, Dr. Eva Zangerle (University of Innsbruck, Austria) on a Context- Aware Music Recommendation project, where I developed and conducted experiments on the #nowplaying-RS dataset containing 11.6 million listening events taken from Twitter and corresponding track features from Spotify. I published my research paper in the Proceedings of the 15th Sound & Music Computing Conference (SMC'18), and gave an oral presentation at Limassol, Cyprus.
Research Intern | May - July, December, 2016
I worked with Prof. Rajat Subhra Chakraborty in the Secured Embedded Architecture Laboratory during summer and with Prof. Rajiv Ranjan Sahay in the Computational Vision Laboratory over the winter.
I was selected to be an intern under the Summer Internship Program of the Computer Science and Engineering department. I worked on Reversible Digital Watermarking, where I used MATLAB and OpenCV (C++) for image processing. I increased time efficiency by 10% by parallelizing code in OpenCV.
Over the winter, I worked on a Deep Learning project for Text and Natural Image Deblurring. I learnt the nuances of Neural Networks and implemented a 3 layer Convolutional Neural Network using the Caffe framework.