Nirmit Desai

As Director of Data and Tools for AI Models at IBM T J Watson Research Center, I focus on delivering enterprise AI models and building tools to optimize the end-to-end model development process, from data curation to debugging AI use case performance. My work is rooted in advanced capabilities in data management and artificial intelligence with a commitment to driving innovation in enterprise AI applications. This work has led to IBM's open-source Granite SLMs, including the latest Granite 4.0, which has an industry-leading performance for its weight class at a fraction of the inference cost.

Previously, I led IBM's Edge Computing research globally, spanning the full-stack edge computing capabilities from edge infrastructures and platforms to enabling data processing and AI at the edge. I have also led the team that created the technology underlying the Flu risk predictions (2020) and Mesh Network Alerts (2018) in the Weather company platforms, and Simulcastr video streaming at the US Tennis Open in 2015.

My work has been recognized with several awards, e.g., NCSU CoE Dissertation Award (2008), INFORMS Innovation in Analytics Award (2012), Best paper award at IEEE SCC (2006), and multiple Outstanding Technical Accomplishment Awards at IBM. In 2017, I was inducted into the Alumni Hall of Fame of the Department of Computer Science at NCSU. My research work is published in 50+ US patents and in 80+ research papers in leading journals and conferences in the area of distributed computing and AI.

Publications: See the DBLP listing or Google scholar.

Patents: See the Google patents page.