Water is an essential part of our lives. One would be remiss to not care for something that is integral to our life on earth; something which has the potential to be nourishing as well as cause extensive damage. As the Director of Machine Learning at Phyn, I strive to minimize the damaging impact of water leaks in homes while building tools to enhance the quality of life and educate people to conserve this precious resource. At this role I’m building a world-class team of machine learning engineers and data scientists who will be revolutionizing how we interact with water by applying a data driven approach. I’m involved with all aspects of machine learning research, planning, development of new technologies and communicating the team’s vision and progress to management.
I’ve previously worked as a Principal Data Scientist and Senior Machine Learning Engineer contributing to the development of algorithms that enable detection of leaks, identification of water usage events and safety monitoring. During this role I’ve championed the adoption of open source tools within the team for data processing, pipelining, management, signal processing, machine learning and data science. My work has led to the development of extensive ground truth databases that enable performance measurement of algorithms. I’ve managed large, nationwide pilots and built tools for visualization that have been used by the company to flag potential issues and communicate progress to partners. Analysis of data has guided the requirements for firmware and hardware teams to integrate sensors in novel ways. My work has led to the adoption of a hybrid edge-cloud approach for processing data.
Prior to joining Belkin, I worked as a research engineer at Recon Dynamics where I was involved in the research and development of algorithms to improve the vertical accuracy in an indoor environment. I’ve also been involved in the development of experimental test beds, management of tools for site survey and calibration during deployment of trial systems. I was responsible for the development of algorithms, data analytics and visualization tools for its asset management solution..
My educational background includes a Doctorate in Electrical Engineering at Clemson University. My primary research was focused on improving indoor positioning to sub-decimeter accuracy using ultra-wideband (UWB) sensors and unaided sensors such as cameras and LADAR using an adaptive filtering framework (Kalman and particle filters). During this time I’ve also been involved in biometrics research using facial features and keystroke analysis, and building test-beds and software to understand the physiological impact of lag from head-mounted displays.
During my free time I enjoy reading books while snuggling up with my cat, traveling, good food and cooking (My wife says I’m a decent cook).
Salil P. Banerjee 