The hardest engineering problems are boundary problems.

Between what teams own and what systems enforce, between what scales today and what breaks next. I work on revenue platforms, domain modeling, org design, and agentic infrastructure.

What I work on

Revenue Platforms

Where product behavior, commercial logic, and financial reality intersect. The work is separating concerns businesses couple by default: what happened, what it's worth, what to charge, how to settle.

Platform Architecture

An ownership model, not a team. Reducing coordination cost so capability compounds and dependency doesn't. DevEx lives where the work lives, not in a rescue squad.

Domain Modeling

Turning business logic that evolved through informal coordination into explicit, enforceable system boundaries. Can this model survive the next three business decisions without a rewrite?

Engineering Organizations

Org design is system design. Team boundaries become API boundaries, and Conway's Law is not a metaphor. The question is always: does this operating model work at 10x?

Agentic Systems

I build my own agent infrastructure. Dispatcher pattern, specialized task agents, shared data store. The interesting constraint is trust: how much autonomy do you give a system that can take action?

Systems Strategy

Strategy without an execution path is decoration. Start from the customer and walk backward through the architecture. If you can't trace a system change to a customer outcome, it doesn't matter yet.

Systems

Not a resume. A view into the types of problems I work on.

Distributed systems

Revenue Platform Architecture

The gap between what a product does and what a business charges for it. Revenue platforms fail when business rules that evolved through informal coordination get encoded as system assumptions. The job is modeling commercial logic as an explicit, evolvable set of domain boundaries.

Domain architecture

Domain Modeling

Turning business logic that evolved through years of informal coordination into explicit, enforceable system boundaries. Most domain models fail not because they're technically wrong, but because they encode assumptions that made sense three business decisions ago.

Platform engineering

Internal Platform Systems

Platform is an ownership model, not a team. The job is reducing coordination cost across teams with very different needs without becoming a dependency.

AI / Automation

Agentic System Design

AngelOS: 50+ autonomous agents across planning, operations, monitoring, and developer tooling. Dispatcher pattern routing user intent to specialized agents. From grocery planning that reads a shared calendar to trading bots with multi-broker interfaces to daily briefing synthesis.

How I think

Prefer systems over fixes

A fix addresses a symptom. A system addresses the class of problem. Always ask: will this solve it next time too?

Make constraints explicit

Hidden constraints cause more damage than hard problems. Name them, write them down, design around them.

Your leaders are your first team

When leaders treat their peer group as their first team, decisions compound. When they don't, every initiative becomes a negotiation between silos.

Reduce coordination cost first

Before optimizing code, optimize how people work together. The highest-leverage improvements are organizational.

Interfaces are organizational decisions

API boundaries reflect team boundaries. Design your interfaces knowing they'll outlast the code behind them.

Say less, show more

Working systems communicate more than slide decks. Ship something small and real over something large and theoretical.

What’s live now

Not a portfolio — a running system. These numbers come from live infrastructure.

6

angel-platform agents

4

Running on this site

7

Active repos

304

Commits (30d)

Chief of Staff Agent

Weekly planning, priority synthesis, and agenda preparation

Weekly

Metrics Agent

Aggregates GitHub stats, deployment activity, and agent health into a unified view

Hourly

GitHub Agent

Monitors repo activity, tracks commit trends, and surfaces relevant changes

Every 30 min

Repo Scanner Agent

Scans all GitHub repos for ADRs, architecture diagrams, and builds a mental model of running systems via OpenRouter

Hourly

Architecture

System design thinking, visualized.

Agent System

Dispatcher pattern for routing user intent to specialized task agents, with shared data store and frontend rendering.

GitHub Metrics Pipeline

Data flow from GitHub API through aggregation service to cached frontend delivery.