Research

I am a Principal Scientist for Alexa Speech Recognition and Deep Learning groups at Amazon. I joined Alexa in 2018 to build the Alexa Pittsburgh organization, which has now grown to over fifty scientists where our mission is to deliver customer-delighting natural language processing experiences through both edge-first and cloud-centric solutions. My primary focus for Alexa has been on low-latency, real-time ML design, and I frequently work cross-discipline with many talented software engineers, product managers, language engineers, hardware architects, legal teams and executive leadership. Generally, my interests lie in the computational aspects of Machine Learning and Artificial Intelligence broadly construed: online learning, distributed learning, multi-task learning, edge machine learning, model compression, efficient data structures for ML and AI under resource constraints, beyond worst case analysis of algorithms, speech and spoken language understanding.

I draw inspiration from different areas, and a central theme of my work is combining the state-of-the-art with classic ideas.

Furthermore, I am a firm believer that an effective researcher within industry should possess a balanced portfolio of skills. Ultimately, for an impactful contribution to society, strong engineering fundamentals are required, accompanied by modern programming practices, design paradigms and new technology utilization.