For business leaders who need AI working in weeks, without hiring a team

I build the agents that build the things you ship.

Staff engineer, fifteen years. Engineering rigor with an artist's finish, on a 48-hour clock. Booked direct, no agency markup, no suits between you and the work.

Work with me → See the work

↳ two live demos sit alongside this hero: an AI-spend dashboard playing back a seeded month, and the agent build pipeline that actually built this site, replaying the real tickets. seeded data, real shape.

Three ways to work together

Pick a shape.

Fixed price wherever possible. The deliverable is always something running, not a slide deck about something running.

most chosen 1–3 weeks

The Build

From $2,500

One problem, one deliverable. Fixed price, no hourly surprises. Shipped to a working URL with a draft PR you can read end-to-end.

Spec a build →

start here 48-hour turnaround

The Intensive

$750

Two focused hours on your operations. You walk away with a written AI roadmap you own. Specific, scoped, opinionated.

Book the Intensive →

long-game ongoing

The Retainer

Ask

Ongoing ops layer. Cancel anytime. Your AI systems stay running, instrumented, and improving. The kind of arrangement where you sleep better.

Discuss the fit →

Selected work

Work, not slides.

Six artifacts you can read about, fork, run, or talk to right now.

case study · 01

lithium

Mood stabilizer for your AI bill.

Local Rust daemon. Polls every provider you pay, normalizes the bill into one SQLite store, answers a single question: what's actually burning this month, fixed and variable, across everything? No SaaS, no telemetry. Just the number.

Read the build → github · 2★

case study · 02

Smithy

Ships software faster by running AI coding agents as a coordinated team.

Fork of OpenAI's Symphony with dual runtimes (Codex + Claude Code), cross-model adversarial pre-PR review, Linear OAuth identity, label-gated autonomous merge, and model-summarized run logs. Same Symphony shape, harness opinions cranked up. v2 in progress.

Read the build → github

case study · 03

Salazar

The tool that builds itself.

Autonomous coding orchestrator. Hand it a markdown spec, planner-generator-evaluator loops until the software exists. Pointed at its own CLI spec: 1,141 tests, four hours, walked away. Pointed at mini-jwt: 38/38 features, 96% coverage, 70 minutes, $9.27.

Read the build → github · 23★

case study · 04

Dealer Tools

An AI support agent that answered customer questions, built before it was easy.

While working at ACME Autos we shipped one of the first AI-enabled customer support agents, before agents were a thing. Tool calling was hand-coded, pre-harnesses, pre-SDKs, pre-MCP. The case study runs a hardened, period-styled version you can talk to. Try to break it. (You're welcome to. It won't.)

Read the build → talk to the demo →

case study · 05

WindWatts

Wind turbine output, mapped over real terrain, in the browser.

Geospatial production-mapping for the National Lab of the Rockies. Full frontend rebuild plus a thin tile-server, deployed at windwatts.nlr.gov. The kind of work where the data has to be right and the UI can't get in the way.

Read the build → github · 4★

case study · 06

Open Brain

Your AI remembers what matters, across every tool you use.

Semantic memory layer for AI agents. Postgres + pgvector + an MCP server. Drop a sentence, it gets embedded, classified, and filed. The next agent that needs the context gets it without being told.

Read the build → github · 9★

Live operations

Three dashboards. Three questions.

Three questions you'll actually ask: what am I spending, is the work getting done, is it staying up. These run my own operation, and the same instrumentation is what I'd build into yours. Mocked here, real endpoint shapes underneath.

lithium /api/ops/lithium

"what's burning this month?"

$453

month-to-date · projected $453

anthropic $106
openai $34
openrouter $13
claude max (fixed) $200
chatgpt pro (fixed) $100

smithy /api/ops/smithy

"what's the harness doing right now?"

1

running

6

merged today

PER-77 merged
PER-78 in-review
PER-79 running
PER-80 queued
PER-81 queued

openclaw /api/ops/openclaw

"are the workers alive?"

198h

uptime · 4 workers

claw-01 · intake idle
claw-02 · research working
claw-03 · compose working
claw-04 · review idle

Whatever you ship, this is the layer that tells you it's still working. Your version tracks your spend, your jobs, your uptime.

~ the smith ~

Who's holding the hammer.

Shawn Petros. Staff engineer, fifteen years deep, operating Petros Industries solo. Shipped for corporate, now shipping for SMB. Building AI tools for smaller teams, on constrained timelines, with no suits between you and the person writing the code.

Based in SoCal. USAF vet. Dad. Stick shift enthusiast. The engineer who never stopped being a creative, and always was a builder.

"Build like a smith.
Ship like a bricklayer."

~ how I work ~

~ the smith, mid-build ~

last 90 days · github

1,573 commits

active

78 / 90

~ proof in the build cadence ~

↳ shipped at

15 years experience/
USAF Veteran/
Atlassian/
Cars.com/
Dept. of Energy/
Northwestern Mutual/
AmFam

~ pull the cord ~

Let's talk.

Tell me what's busted and how soon you need it not-busted. I'll write back with a recommendation, not a slide deck.