−0%
monthly cloud spend
Migrated an EC2 monolith to AWS serverless microservices for a US data startup.
Córdoba, Argentina · --:-- UTC−3
I'm Jorge, a senior full stack engineer who turns messy requirements into production systems people rely on. Node.js and React on top, AWS, Azure or GCP underneath, and one person accountable for all of it.
Example workflow: write a spec, run tests, deploy to production with zero downtime, measure a 15% lift in purchase completion.
Spec-driven workflow
Claude Code · Codex · tests first
Production code shipped for
Not story points. Not lines of code. The metrics a CFO or a head of product would put on a slide.
−0%
monthly cloud spend
Migrated an EC2 monolith to AWS serverless microservices for a US data startup.
+0%
game purchase completion
Fallback path for network and payment-provider failures at checkout.
~0%
faster initial load
Main purchase flow, rebuilt mobile-first.
0
microfrontends upgraded, 0 downtime
Major versions of Node.js, React and Ant Design.
~0%
test coverage, from zero
Unit + end-to-end on critical paths.
0 mo
blank repo → live SaaS
Angular SaaS for a fintech serving millions.
Hiring me means one person owns the outcome: architecture, backend, frontend, cloud and delivery. Fewer meetings, fewer gaps, faster answers.
Node.js APIs and React or Angular frontends, from Figma to production. I design the system, build both sides of it and keep it running.
Serverless, microservices and the boring-but-critical bits: config, secrets, scheduled jobs. Designed to scale — and to shrink the bill.
Monolith to microservices. Major-version upgrades across dozens of modules. Microfrontends. All while the product keeps shipping.
I work with Claude Code and Codex, guided by written specs and real tests. You get the speed of AI with the guardrails of a senior engineer.
The problem — Purchases on the store were dropping whenever the network or a payment provider hiccupped — lost revenue on every failure.
+15%
purchase completion
~35%
faster initial load
Stack
Case 01 / 04
The problem — A growing go-to-market data platform was locked inside a single EC2 monolith: expensive to run, risky to deploy, hard to scale.
−83%
monthly cloud spend
~80%
coverage on critical paths
Stack
Case 02 / 04
The problem — Fifteen microfrontends pinned to aging major versions of Node.js, React and Ant Design — and a roadmap that couldn't pause for a rewrite.
15
modules migrated
0
delivery downtime
Stack
Case 03 / 04
The problem — A national-scale fintech needed a new product live fast — without compromising the reliability its users expect from their money.
5 mo
zero to production
M+
users on the platform
Stack
Case 04 / 04
We define what success means in numbers — revenue, cost, speed — before anyone opens an editor.
Written specs and acceptance criteria. The contract between us, the tests and the AI tooling.
Backend, frontend and infra in small, tested increments. You see working software every week.
Ship, measure, iterate. Docs and handover included — and I stay accountable after launch.
Picked for the problem — not the hype.
Backend
Frontend
Cloud
Quality
AI
Apps I designed, built and shipped for my own clients — from store listing to backend.
I started in 2018 writing backend services at IBM — and, at the same time, co-founding my own product company. Doing both taught me early that code only matters when it moves the business.
Since then I've built finance products used by millions in Argentina, re-architected a US data startup's cloud, and kept the Xbox store's checkout standing when the network wasn't. Today I work independently with companies that want a senior engineer who owns the outcome, not just the ticket.