Machine Intellection

Core Algorithmic research in machine learning, deep learning and beyond.
Research thrusts include learning with 10x fewer labeled samples, incorporating knowledge graphs into deep learning, online learning, white-box learning, transfer learning, data driven optimization, and deep learning on encrypted data. Beyond deep learning, the team is beginning to explore long term research areas like common sense reasoning and quantum ML.

Focused on serving the wide ranging AI needs from MNCs to government agencies. Verticals of interest range from personalized medicine, finance, health-care, IoT, aviation, transportation, advanced semiconductor manufacturing, and more.

We are looking beyond GPUs and TPUs, and reimagining the entire hardware stack for deep learning from algorithms all the way down to silicon. The R&D roadmap consists of SRAM accelerator chips, in the short term, all the way to monolithic 3D deep learning systems in the long run.