I have been building software professionally since 2000, starting as the primary developer on an MLS real estate platform at an early-stage startup. The industries have changed since then (travel, enterprise network security, online gaming), but the work has stayed consistent: backend systems at scale, teams worth building, and technical problems worth solving properly.

My career has run through acquisitions, startups that became something bigger, and organizations where I was both the technical lead and the people manager. The resume has the full timeline. What it cannot say is that the thing keeping me in this field for 25 years is the same as it was in year one: a pull toward systems that are just complex enough to require real understanding before you can improve them.

How I Lead

The roles I am drawn to sit at the intersection of management and technical depth. Not pure people management, not pure IC. The most useful place to be is where you can set technical direction, unblock teams with hands-on POC work, and grow engineers who will eventually not need you for the problems you helped them solve.

I have grown teams from scratch and owned the full people-management stack along the way: hiring, 1:1s, performance reviews, promotion planning, and compensation input. The headcount was never the interesting part. Watching a junior engineer crack a hard problem because you gave them the right context and then got out of the way is the part worth doing again.

Engineering effectiveness is the other thread I keep pulling on. Find the friction engineers are fighting, remove it, and make sure the next person does not have to fight the same battle. Better tooling, cleaner service boundaries, faster paths from code to production. The details change. The goal does not.

How I Learn

The clearest example is the Complect compiler. I had no background in compiler theory. I wanted to give a technically substantive conference talk, so I built a working compiler from scratch in Node.js, taught myself the theory along the way, and shipped it in time for OpenJS World 2022. The talk landed. Then the project kept going.

I later added an LLVM backend so the language compiles to native binaries via clang, then SDL2 integration for 2D and 3D graphics. Complect is now a personal research platform for compiler techniques and low-level systems work. This is the pattern I keep coming back to: pick a real constraint, build something that actually works, and do not stop when the deadline passes.

What I Am Working on Now

Three ongoing projects. A C++ high-performance tokenizer targeting SIMD and AVX2 optimization for LLM training at terabyte scale. A research platform for evaluating RAG document chunking strategies, with a novel NLP-based approach and a paper as the target output. And EDU, an autonomous multi-agent AI learning system designed to replace grade-based education with mastery-driven, personalized learning for K-12 students.

These are long bets. Each one is worth writing about, and I intend to.

Beyond Code

From 2017 to 2022 I served on the City of Richmond Child Care Development Advisory Committee, eventually as Chairperson. The work included advising City Council on child care planning policy, reviewing grant applications, evaluating facility operators, and contributing to the Richmond Child Care Needs Assessment, a five-year civic plan adopted by Council. From 2023 to 2025 I chaired the Homma Elementary Parent Advisory Council.

Good leadership is the same problem whether you are running an engineering org or a school PAC. Get the right people aligned, make decisions with incomplete information, and be honest about what you do not know yet.

Find Me

GitHub  ·  LinkedIn  ·  Twitter/X  ·  ORCID  ·  Google Scholar