How I think, lead and ship.
Principles, technical leadership, AI workflow and live GitHub activity: the detail behind what you see on the home page.
How I build software.
Architecture before code.
I draw the system before I open the editor: boundaries, data flow, failure modes, blast radius. Architecture isn't about elegance — it's about making the next change cheap.
Built to be maintained.
The diagram is the starting point, not the deliverable. I turn that architecture into code the team can read, extend, and change without fear — less accidental complexity, more explicit decisions.
Performance is a feature.
p95 is a budget, not a vanity metric. Nobody cares about Lighthouse screenshots — they care about fast interfaces, stable flows, and software that doesn't break under pressure. I measure first, and I won't call something done if it regresses TTFB or hydration.
AI-assisted engineering.
AI accelerates the repetitive, the operational, scaffolding, and large refactors. The architecture, the tests, and the production cut stay the engineer's call. Nothing reaches production unchecked.
Clarity over cleverness.
The smartest solution nobody else understands is a liability. I prefer code and documentation the next developer — or the next AI tool — can safely pick up.
Good software at scale is built on shared understanding.
Senior engineering is half writing software and half making the team that writes it better. No heroics, no single-person bottlenecks.
Alignment over heroics.
Documented decisions and shared context so the team moves without depending on one person. Continuity shouldn't live in anyone's head.
Architecture you can understand.
I write decision records the next engineer reads in fifteen minutes — and that AI tools can follow. Shorter onboarding, less friction adding people.
Reviews that make better engineers.
I review for correctness and readability, but mostly so the team writes better in six months than it does today. Code review is knowledge transfer, not a formality.
Security and sustainable delivery.
I've worked in environments with security and process requirements without slowing delivery. Secure practices as part of how the code is written — not a checklist taped to the end.
Engineers in the loop. AI on the rails.
I use AI assistants in every project to speed up the repetitive, the operational, and large refactors — scaffolding, migrations, documentation. The architecture, the tests, and the production cut are the engineer's call. The difference is throughput, not judgment: nothing reaches production unchecked.
- Claude
- Codex
- Copilot
- JetBrains AI
- Gemini
- June AI
Scaffolding & migrations.
AI generates the scaffolding and the repetitive changes; I review, type-check, and commit.
Large refactors, with judgment.
It accelerates heavy refactors and migrations; the design decisions and boundaries are mine.
Documentation as I go.
Every architecture decision gets a one-paragraph ADR — drafted with AI, signed off by me.
Always shipping.
Live numbers from GitHub.
Let's talk about what you're building.
Whether you need a technical partner, senior engineering leadership, or simply someone to ship the product right — for a startup, an e-commerce brand, a product team, or an agency — start here.
Message sent. I'll reply about your project as soon as possible.
Message sent. I'll reply about the role as soon as possible.
Couldn't send. Try again or email matt.bubello@gmail.com directly.
Or email directlyBuenos Aires (UTC−3)