For US and EU founders and engineering managers
Your roadmap is growing faster than the team that ships it.
Bring in a senior engineer on contract to take one hard piece to production: a new AI feature, a system that keeps breaking, or the senior seat your team is missing. In English, async, documented.
After 20+ years shipping software, I know which piece to ship first and which one only creates maintenance, so you do not pay for that lesson on your budget.
What stalls right now
The feature that matters keeps sliding a sprint. A service breaks in production and nobody owns it end to end. You need a senior pair of hands now, not a hire three months out.
Proof of shipped work
You can read the work before you trust it.
I have written production Rails since 2012 and shipped software for 20+ years, including a platform that serves around 1.8 million visits a month. The tools I use to build with agents are open, so you are not taking my word for how I work: you can read the code.
- PlanQueue
The queue that drives my own agents. You can read exactly how the loop, the retries and the human approval work.
- claude-code-harness
The agent harness I run every day: evals on real examples, a human in the loop, the smallest useful workflow first.
- serpapi-agent-toolkit
Tooling I built to give agents reliable search, open so you can see the failure handling, not just the happy path.
- Tapa
A shipped side project, public end to end, so you can judge how I structure and finish real code.
Ways to work together
Ship an AI feature to production
You have an AI feature on the roadmap that keeps stalling between a good demo and something users can rely on. I take it the rest of the way: evals on your real examples, a human in the loop where the risk is, and the smallest useful version live first, so you see it working before you scale it.
Rescue a production system
A service keeps breaking, deploys make you nervous, and the person who understood it is gone. I stabilize it, cut the release risk, and hand it back with the failure modes written down, so the next incident does not depend on one person's memory.
Fractional senior engineer
Your team is good but short one senior seat, and a full-time hire is months away. I plug in part-time on Slack and GitHub: the hard reviews, the architecture calls, the piece nobody has time to own. You get senior judgment now, and the decisions stay documented instead of in one head.
Case studies
Result first, then how it was done.
Result: about 30% lower cost and 40% less time to ship new versions of a live system.
Situation: every release was slow and manual, the team burned hours checking by hand, and each deploy carried real risk. Action: I mapped which checks could be automated and kept the risk decisions with the right person, instead of automating blindly. Result: faster releases without the fire drills.
Result: a platform serving about 1.8 million visits a month, kept stable in production.
Situation: real traffic, real uptime expectations, no room for a fragile setup. Action: I owned the parts that had to hold up under load and made the operational limits explicit. Result: it kept running under production traffic, which is the only test that counts.
Result: agents I run in production every day, open for anyone to read line by line.
Situation: most AI work is a screenshot you cannot verify. Action: I built and operate my own agent tooling in the open, with evals and human approval baked in. Result: before you trust me with your system, you can inspect exactly how I run one.
How we work
A contractor in another country only works if the collaboration does.
- Based in Brazil (UTC-3), I overlap your full US morning, so you get same-day answers, not next-day.
- Written progress before your standup, so you always know where the work stands without chasing me.
- Decisions get documented as we go, so nothing important lives only in my head.
- The work is inspectable before you commit: we start with a small, paid, fixed-scope engagement, so a wrong bet costs a scoping call, not a quarter.
Taking contract engineering work with companies in the US and Europe, in English.
The first step is small and concrete.
Book a scoping call. We look at the one piece that matters most right now, and you leave with a concrete next step, whether or not you hire me to build it.