Led 400+ engineers and product managers focused on AI development and adoption for J.P. Morgan Corporate & Investment BankStrategy / Product / Architecture / Engineering / Delivery / Adoption / Research

Co-founded J.P. Morgan AI Research, which conducts core research in areas of AI that are of particular interest in financial services. https://www.jpmorgan.com/technology/artificial-intelligencePublished novel research at prestigious venues including NeurIPS, ICML, AAAI, IJCAI and ICAIF. Topics include game theory, multi-agent reinforcement learning, large-scale simulation of ML agents, link prediction in multiplex networks, synthetic data generation, and textual insight generation from financial data.

R&D for ML on Android/IoT and Nest Home Intelligence. Developed AI algorithms using behavioral game theory and reinforcement learning-based control techniques for robotics products; fusing data from ambient sensors in IoT products and inferring human behavioral patterns using machine learning. Managed the deployment of Choreograph for IoT orchestration in Google Assistant. Led the development of ground-breaking computer vision and robotic articulation showcases using Android on Raspberry Pi -- this work was featured as a centerpiece of the Android display at Google I/O 2017 and was then enhanced to become a permanent display in the main lobby of Google HQ! https://android-developers.googleblog.com/2018/05/all-android-things-at-google-io.htmlhttps://www.raspberrypi.com/news/android-things-expression-flower/

Led the development and launch of the Home/Away Assist features in the Nest ecosystem, bringing together all of Nest's products into an integrated ecosystem in a novel cloud-based machine learning architecture. I did the initial hands-on ML research and later led the multi-year product development and launch all the way to deployment to over 40 million homes.

Founded a tech startup that provided AI-powered automated services to help residential consumers navigate deregulated retail electricity markets. https://www.newscientist.com/article/mg21929315-200-ai-systems-switch-your-energy-bills-to-save-you-money/We collected historical data across US utilities and applied machine learning and game theory to continually figure out the best match between consumers and retail electricity plans.In addition to the strategic, financing, and recruiting responsibilities of the CEO role, I also performed key CTO/engineering and product roles in defining how the underlying technology should be productized and marketed.We won the McGinnis Venture Competition hosted by Carnegie Mellon's Tepper School of Business and were funded by the Open Field Entrepreneur's Fund.

Developed machine learning and AI algorithms for distributed multi-agent control with applications in cooperative mobile robots, wireless sensor networks, and Smart Grid agents.Developed novel learning-based localization and motion planning algorithms for human-robot sensor networks as part of the DARPA LANdroids project. https://link.springer.com/chapter/10.1007/978-3-642-25832-9_76 Co-created Power TAC, an open-source agent-based simulation environment for distributed Smart Grid research and competition, with collaborators in the US and Europe. https://powertac.org/

Promoted rapidly to Managing Director through multiple highly technical software engineering and engineering management roles.Developed the MSNet C++ and Java software libraries for networked parallel programming and reliable high-speed messaging; these libraries form the backbone of most electronic trading systems within Morgan Stanley to this day. Also led the architecture and engineering of scalable middleware for pub-sub, service discovery, entitlements management, and grid computing. Proposed, created and recruited a core team of ~15 software engineers to manage the entire Java framework and middleware infrastructure for 1500+ software developers across Morgan Stanley. Changed roles to lead software development teams for systems that managed algorithmic trade execution (multi-benchmark agents, smart order routing, dark liquidity pools) and exotic equity derivatives trading. Managed a team of ~40 globally dispersed software engineers.Returned to core technology as Global Head of Application Infrastructure to manage ~120 software engineers responsible for all the software infrastructure including toolkits and middleware across C++/Java/.NET/Python and drove new initiatives to provide software infrastructure for scalable time-series analysis and backtesting of algorithmic trading strategies. Defined product scope and requirements and managed cross-functional project execution in a product management role in addition to the primary engineering management role. Worked with clients within the Technology division and across each business unit to manage product priorities, strategic direction, management communication, and internal product marketing.Initially participated in and later chaired the ArcDev technology architecture committee, which reported directly to the CIO/CTO. The committee led initiatives to mandate and/or recommend technology strategy (e.g., SOA) for all software development within Morgan Stanley.