I investigate failures no one fully understands, find the problem underneath the problem, and turn what I learn into durable products, platforms, and engineering practices.
I am an engineering leader and principal architect with more than 20 years across distributed systems, payments, data platforms, security, and AI/ML. I work at whatever altitude the problem requires, from production code and experiments to architecture and organizational design.
Teams founded from scratch and organizations scaled past 60 · technical leadership across organizations of 200+ · named inventor on 20+ patents
Based in San Francisco.
I trace real system behavior, separate causes from symptoms, and fix the architecture and practices that allowed the failure to persist.
I enjoy taking products and technical capabilities from an ambiguous beginning through implementation and production.
I want the solution to survive through better tests, tooling, observability, ownership, and organizational design.
Understand the system → fix what is broken → prevent recurrence → build the better version
The work is done when it exists, works, and holds.
I set architecture for critical payments systems moving approximately $3 billion annually, spanning a rebuild of the ACH and debit-card stack across cloud, database, and service architecture.
It is a return to a problem I worked on years earlier. I am a named inventor on granted patents covering transaction and location verification.
Before that, I built and led the central data and ML platform at NinjaTrader, a Kraken company, cutting data latency from roughly ten hours to twenty minutes and shipping a customer-facing AI product months ahead of roadmap.
I founded Rivian’s AI and ML research team and worked across vehicle charging, battery systems, factory robotics, and fleet data.
Reinforcement learning reduced DC fast-charge time by 35 percent, while ML root-cause analysis across factory robotics saved more than $3 million annually.
I led research across mobile networks, extended reality, distributed computing, and privacy, directing an $8 million portfolio and collaborations with MIT and other partners.
The work included systems that prove physical presence without disclosing coordinates, analyze distributed sensor data without exposing it, and place computation across devices, edge infrastructure, and networks.
I was named inventor on more than 15 patents and co-authored DeepCompress, which reduced encoder convolution operations by 8 percent and parameters by 20 percent while matching baseline quality. Replication code is on GitHub.
I also led nearly 400 Ericsson volunteers in a COVID-19 response effort recognized by the Global Business Alliance as its best corporate initiative.
I joined Wikipedia during its early growth, helped scale the site to hundreds of servers and more than 50 million daily pageviews, and built one of its earliest automated anti-abuse systems.
That work established a pattern that has followed me throughout my career: understand emerging system behavior, build the immediate intervention, and leave behind a durable capability.
I am a named inventor on more than 20 granted and pending patents across AI/ML, privacy, security, extended reality, electric vehicles, battery systems, and distributed infrastructure.
My research often begins where digital systems meet the physical world:
How can a device prove presence without reporting its location?
How can a network answer a question without exposing the underlying data?
How can a physical system detect when its internal model has drifted from reality?
How should computation be divided among devices, vehicles, edge systems, and networks?
Before returning to industry full time, I conducted research at the San Diego Supercomputer Center and Emory University in causal inference, sparse data, privacy-preserving extraction, and high-performance computing.
From 2020 to 2021, I served as an inaugural Tech Fellow at the Tony Blair Institute for Global Change, working on AI and energy-infrastructure resilience.
I write about engineering systems, AI, reliability, organizational decision-making, and the things I wish someone had written down sooner.
My path into technology was unconventional. I left high school without graduating and earned a GED, then later studied at Columbia and George Washington and completed a PhD in computational social science at UC San Diego.
That experience made me skeptical of conventional signals of potential and attentive to the systems that determine who receives opportunity, mentorship, and credibility.
My background in visual art continues to shape how I think about structure, composition, legibility, and the relationship between form and function.
Outside work, I spend a great deal of time walking, whether across San Francisco or during weeks of hiking and camping in Japan. Walking is how I observe places, understand how they fit together, and notice the systems people build without consciously deciding to build them.
I maintain close ties to both my American and New Zealand roots and serve on the board of the New Zealand American Association of San Francisco.
Have a difficult problem?
I am most interested in consequential systems where the failure is real, the answer is not obvious, and solving the immediate problem is only the beginning.
Contact me →