Rishabh Patra

I am a Software Developer at Amazon Appstore under the Devices and Services organization of the company. My team works on (and maintains) the Developer Experience pipeline, which is a collection of microservices that power ingestion of apps submitted on developer.amazon.com, the app approval process orchestration, and push app metadata and binaries to the FireTV catalog. Our team also handles the distribution of binaries to FireTV devices, where we use a combination of business rules and technical compatibility evaluations to provide binaries to devices based on the software and hardware capabilities of the particular device.

Prior to joining Amazon, I completed my undergraduate at BITS Pilani, Goa Campus, India with a Bachelor of Engineering (B.E.) (Hons.) in Electrical and Communications Engineering with a Minor in Physics.

I had the privilige of working with Ramya Hebbalaguppe, Dr. Tirtharaj Dash, Dr. Ashwin Srinivasan, and other researchers from TCS Research Labs during my final year of undergraduate studies. Our group worked on Confidence calibration of neural networks. Confideence calibration implies the predicted probabilities of the model are representative of the true likelihood. Our group was specifically interested in calibration techniques that maximise the number of high confidence in-distribution samples, while not sacrificing IND accuracy and OOD detection performance. Our research here was published at WACV 2023 as a spotlight paper.

Prior to this, I had interned at TCS Research, where our group worked on explainable methods for Deep Learning. My research here looked at methods to explain the predictions of Deep Learning models using rule based engines combined with Deep Neural Networks on the medical imaging problem of Arrhythmia detection. Before this, I had also worked as a Research Fellow at Maritime Research Centre, Pune. Here, I was exploring methods to model underwater acoustic propagation in the tropical littoral waters in the Indian Ocean Region using Machine Learning methods.

I was also associated with the Anuradha and Prashanth Palakurthi Centre for Artificial Intelligence Research (APPCAIR) in the final years of my undergraduate studies. APPCAIR offered me oppurtunities to work on industrial research projects. I had initially worked on Paper-web break detection (system fault detection) and interpretability of trained neural networks in the domain. APPCAIR helped me secure internships at TCS Research Labs, and the WACV publication was an outcome of the collaboration with APPCAIR.

View: Curriculum Vitae