What I Build

Builder First

I’m a builder, not a consultant. The products are the work.

Everything I know about AI-augmented development, personal data infrastructure, and orchestrating agents across a full portfolio comes from shipping these products day after day — not from advising about them from the sidelines. This page is the evidence.

The Portfolio

Tractor & Silo — privacy-first life-tracking ecosystem

The flagship. A platform across web, iOS, and macOS for the kind of personal data a generation of us has been deliberately capturing for years without anywhere worth keeping it. Built around a capture–enrich–synthesize arc: capture and enrichment ship today, cross-source synthesis is what we’re building toward.

Solo-built and operated since 2016. Rails web app, native iOS app shipped to the App Store, macOS companion via Lens. See the ecosystem →

Lens — quantified-self desktop companion for macOS

A local-first macOS app that captures what happens on your computer — screen, audio, context — and turns it into hourly recaps and daily summaries you can read back. Phase 1 ships standalone today; later phases push journals into Tractor & Silo and pull the timeline back, with cross-source synthesis as the longer-term destination. The desktop half of the ecosystem.

PKM Vault — personal knowledge management with LLM enrichment

A document ingestion pipeline that pulls anything I read, write, or annotate into a structured vault. Five-Ws metadata extraction, local-model enrichment via LM Studio, and a retrieval layer designed to work with the rest of the portfolio. The first production test of an always-on local agent inside my own stack.

What the Practice Looks Like

The same AI-augmented workflow that runs my day job as VP of Engineering also lets me operate the entire portfolio end-to-end as a solo developer. The capabilities below are visible in the products themselves — they’re not pitched, they’re demonstrated:

• AI-native operations — orchestrating agents across a full product portfolio, from code generation to autonomous review cycles to long-running planning workflows that span multiple repos
• Personal data infrastructure — designing systems around the capture–enrich–synthesize arc, with the person using the product owning the substrate, on devices they control
• Technical strategy & architecture — system design, data modeling, and infrastructure decisions for products you intend to live inside for decades, not quarters
• Engineering leadership — code review standards, workflow patterns, and the discipline that lets a one-person engineering team move like a much larger one

A note on consulting

I’m not actively taking on consulting. The portfolio is the work, and that’s where my time goes.

If you’re working on something where this kind of building genuinely fits — AI-augmented development practice, personal data infrastructure, or the long-form question of what software you’d want to still be running in twenty years — I’m reachable. No pitch, no engagement model. Just a conversation if it’s the right shape.