Stealth mode

Heading all Analytics and Digital efforts.

Heading a team to integrate data and computational analytics into strategic decision-making processes across Sumitovant and all Dainippon Sumitomo subsidiaries through the DrugOme, the computational ecosystem developed by my team.

Integral role in raising more than $4B for Roivant. Built the team that created the DrugOme -- Roivant's proprietary analytic engine for drug in-licensing. The DrugOme integrates data from across the landscape of pharma to enable focused analytics to support drug valuation, fine-grained epidemiology, clinical and pre-clinical strategy optimization, and commercial strategy among other analytic activities. Activities on the Computational Research team include algorithmic epidemiology on Real World Evidence data sources, healthcare-specialized Natural Language Processing using a proprietary NLP stack designed in-house, machine-learning driven drug target and toxicology prediction, and drug valuation modeling analytics.

Manager/technical lead of functional reliability team responsible for functional reliability qualification of 14nm and 7nm semiconductor process technologies. Responsible for the development of infrastructure and techniques for screen, wafer/package test, stress, and data analysis of test circuits to meet technology qualification milestones to schedule. Research projects include: developing techniques to measure > 6 sigma transistor variability, comprehending the impact of stress and variability on intrinsic reliability, assessing the impact of AC stresses on realistic circuits, and identifying/modeling extrinsic failure events. Worked in technology consortium to enable and qualify 32/28/22/20nm high-k metal gate CMOS technologies.

Working on technology development for 45nm generation of NOR flash. Developed measurement techniques and software for characterizing flash cell reliability. Developed physics-based reliability models for lifetime estimation and prediction and to allow cell structure optimization for reliability hardness. Conducted various research projects to understand and improve device hardness to random telegraph noise, stress-induced leakage current, negative bias temperature instability, 1/f noise, hot electron stress, and various disturb mechanisms. Worked on breakdown and charge trapping characterization of oxide/nitride stacks, developed software for automated implementation of experiments and analysis of data both at single device and full array levels.

See Numonyx summary

Principle Investigator on Project “Atomic Layer Epitaxial Growth of Half-Metallic and Dilute Magnetic Semiconducting Nanostructures”. Project’s focus is utilization of Chemical Vapor Deposition and various lithographic techniques to enhance the response of sub-100nm scale Giant Magneto-Resistive structures using novel materials, and the development of Ferromagnetic Semiconductors with a Curie Temperature above room temperature. Simultaneously involved in semiclassical simulation of nano-oxide layer spin valves and simulation of noise in current-perpendicular to the plane spin valves. Initiated study of ballistic transport in giant magnetostrictive nanowires.

Redesigned and expanded an equation compiler for a materials database to interface with Nanoelectronic Modeling Program (NEMO) using yacc. NEMO simulates devices that exhibit quantum effects such as RTDs.

Designed and programmed an application to visualize historical data of Britain using 3D graphics. Application programmed in C++ with Open Inventor on SGI workstation.