PRINCIPAL SOFTWARE ENGINEER

Built to scale. Designed to operate.

15+ years building high-stakes backend systems across fintech, commerce, and real-time platforms: from software managing trillions in assets to production ML whose results were highlighted on an earnings call.

The constant across all of it: improving the quality of life and efficiency of the people around me.

  • I reduced release cycles at First Republic from days of coordination to minutes, for hundreds of engineers.
  • I led and completed the initiative at Cimpress that freed multiple teams of senior staff from brownfield maintenance to build products that add value to the company.
  • I created a startup, designed its system, and led the team that reduced development and hosting costs by 90% for the companies we worked with.

Highlighted engineering work

Architecture matters when it changes what an organization can ship, operate, or afford. These examples come from paid engineering roles.

CarGurus · Principal Software Engineer

Production ML shipped by a backend team

Machine learning was the exclusive domain of the data science team. I led the first non-data-science team to ship a production ML application with a full MLOps pipeline; it was highlighted to investors on an earnings call, and the ML training I ran made that path repeatable for other backend teams.

First Republic Bank · Senior Software Engineer

Enterprise delivery platform

Deployments took days, sometimes more than a week. I built and presented the CI/CD platform the bank adopted, then trained teams on how to build on it. Company-wide deployments moved from days to minutes on a system used by hundreds of engineers.

Cimpress / Vistaprint · Lead Software Engineer

High-scale commerce rebuild

Tax-lookup effort multiplied with each acquisition, heading toward a failure state estimated at $15 million a week in revenue. I won buy-in for a rebuild, automated the human-heavy steps, and extracted cached, horizontally scalable microservices that a small junior team ran smoothly at thousands of queries per second.

Apocoplay · Founder and CTO

Multiplayer infrastructure economics

Server-side development kept multiplayer games out of reach for most studios. I designed and built an event-stream engine with routing and filtering while leading a 20-person team; it cut server and development costs by 90% per game, saving client studios hundreds of thousands of dollars per title annually.

Featured technical proof

Black Skies: a distributed-systems design built to be disproved.

Black Skies is a persistent multiplayer space game designed for 10,000-player battles with no time dilation. The game is the setting; the engineering questions are general: bounded consistency, ownership and fencing, failure recovery, delivery fanout, benchmark design, and explicit falsifiers. The series separates what is implemented, what is measured, and what is still a hypothesis.

  • IMPLEMENTEDThe coordination machinery exists as runnable code: tile lifecycle, epoch-fenced ownership, and the Redis commit path, behind 1,800+ automated tests and chaos suites.
  • MEASUREDSingle-tile load testing at full saturation has held its tick budget, and local benchmark implementations exist for every ledger row.
  • PENDINGFour production gates stand open with no evidence attached: live Redis Cluster, live Aurora Limitless, the access-boundary proof, and benchmark rows at scale. Until they close, 10,000 concurrent players in one battle is a pre-registered design target with published falsifiers, not a measured result.

Current independent work

CogniCluster is my current independent engineering lab. PlotWarden is the durable orchestration system; Black Skies is the distributed-systems proving ground and product it helps build. PlotWarden is a working prototype with a published specification and kill condition.

Looking for a Principal engineer?

Available for full-time Principal and Staff-equivalent engineering roles in distributed systems, platform engineering, production ML, and AI infrastructure. Boston area or remote.