About

Sandeep Singh Sandha is a highly skilled Ph.D. candidate in computer science at UCLA with a Bachelor's degree in computer science from IIT Roorkee. His research interests lie in designing robust production Machine Learning/Reinforcement Learning pipelines in the presence of runtime uncertainties. He is excited about the use-cases in the domains of cyber-physical systems, decision-making systems, databases, smart cities, autonomous vehicles, and mobile computing. At UCLA, Sandeep leads the CS-GSA (Graduate student association of CS students) and STEM Summer Program. He is also a research collaborator with Amazon AI, Arm Research, and IBM Research. His work on parallel hyperparameter tuning at Arm is open-sourced and currently used in production. Mango, the tool he developed, is available on GitHub. Sandeep also has experience working as a Research and Development Engineer at IBM, where he contributed to the development of several projects. He formulated complex parallel and distributed algorithms that solve partial differential equations describing the physical body's behavior. He also worked on real-time and historical water-quality sensing, analytics, and data integration, where he developed an Internet of things framework from scratch, with the help of mobile computing, Cloud Computing, and Sensor networks to do water quality data collection, analytics, and user engagement in real-time. Sandeep's past experience includes a research internship at Teradata, where he worked on enabling data-parallel training within the database architecture. He implemented distributed stochastic gradient descent, database abstractions, and demonstrated its usage to train traditional models along with the state-of-the-art deep neural networks. This work received a lot of interesting reactions from the database community. Sandeep's technical skills include Research Scientist, ML, Data Scientist, Reinforcement Learning, Hyperparameter, AI, Mobile, PHP, Neural Networks, Integration, dotNET, and NoSQL. He has 4.80 years of relevant experience.