~$gabor/ lorenzo

~$ whoami

Full-stack engineer. I ship AI-built software that survives real users.

I'm Lorenzo. I build and ship production web products, working with AI agents every day and keeping my own judgment on every merge. I have worked in a long-lived legacy codebase, where I learned to respect the consequences of every change. The rest of the time I build my own products, plus the tests that keep them honest.

See what I've shipped →Get in touch →
  • AI Product Engineer
  • Applied AI Engineer
  • Forward Deployed Engineer
  • Full-Stack Engineer
  • Product Engineer

TypeScript · Next.js · React · Supabase · PostgreSQL · Stripe · Vercel

Open to engineering roles and one freelance sprint

# for teams

I don't hand over code. I hand over software that works.

I design, build and ship an entire product alone: from login to payment to the checks that keep a release standing. Inside a team I bring the same end-to-end ownership.

# for founders

One broken flow, fixed in a focused sprint.

Authentication, onboarding, Stripe payments or deployment. Reproduced, fixed, covered by regression tests and delivered as a review-ready pull request.

~$ ls ./work

4 items

ProposalPilotLIVE SAASA SaaS I designed, built and shipped alone: authentication, onboarding, AI proposal generation, client sharing, Stripe billing and release gates. It runs in production today.Next.js · Supabase · Stripe · TypeScript · VercelmicroGPT in COPEN SOURCEA full GPT written in pure C, no PyTorch and no runtime dependencies: tokenization, attention, RMS norm, MLP, backpropagation, Adam, generation, save/load and a training CLI. Built to understand the machine below the prompt box.C · Transformers · zero dependenciesOpenX GrowthPUBLIC ALPHAA self-hosted workspace for growing on X without handing your credentials to a third party: official APIs, OAuth 2.0 with PKCE, encrypted tokens, drafts, scheduling, analytics, a REST API and a local MCP server.TypeScript · Cloudflare Workers · D1 · OAuth 2.0Doclify GuardrailOPEN SOURCEA small local-first CLI for repositories where people and AI agents edit the same markdown. It catches documentation drift before it ships.Node.js · CLI · local-first

~$ cat ./production-fix-sprint

The Production Fix Sprint

A Production Fix Sprint is a fixed-scope engagement for one critical flow in an AI-built SaaS. I reproduce the failure, ship the production-safe fix, add automated regression coverage and hand over a review-ready pull request with the remaining risks written down. Duration depends on complexity and is agreed after an initial review of the flow and codebase.

  • Authentication
  • Onboarding
  • Payments
  • Deployment
one flowscope, duration and price agreed before we start
Tell me what's broken →
  1. 01

    Reproduce

    "It breaks sometimes" becomes a failing path we can run on demand.

  2. 02

    Fix

    The smallest production-safe change, without turning the sprint into a rewrite.

  3. 03

    Prove

    Automated coverage for the failure, plus the checks that matter around it.

  4. 04

    Hand off

    A review-ready pull request, the decision trail and every remaining risk in plain English.

~$ cat ./about.md

I use AI. I keep the judgment.

Lorenzo is a product-minded full-stack engineer based in Italy. He builds web products with TypeScript, Next.js, React, Supabase, PostgreSQL, Stripe and Vercel, and works with AI coding agents as part of his daily workflow.

He has worked in a long-lived legacy codebase, where he learned to respect blast radius and the operational consequences of each change. Outside of it he designs and ships his own SaaS products end to end, from authentication to billing to release gates.

That combination is his edge: old software teaches blast radius, AI-built software teaches how fast a plausible change can slip through. So he reproduces the failure, narrows the fix, makes the proof runnable and writes down what still carries risk.

LorenzoItaly · working worldwide

~$ ./faq

8 answers

What do you do?

I am a full-stack engineer. I build and harden production web applications with TypeScript, Next.js, React, Supabase, PostgreSQL and Stripe, and I work with AI coding agents as part of the process.

What is a Production Fix Sprint?

A fixed-scope engagement on one critical SaaS flow. The failure is reproduced, fixed, covered by automated tests and handed over as a review-ready pull request with the remaining risks documented. Duration is agreed after reviewing the flow, code and complexity.

How does pricing work?

Every engagement is quoted after a short look at the flow and the codebase, and the scope is fixed before we start. Longer engagements and full-time roles are discussed case by case.

Which flows do you fix?

Authentication, onboarding, Stripe payments and deployment: the paths a user cannot skip and a release cannot ignore.

Are you available for full-time roles?

Yes. I am open to product engineer and full-stack engineer roles, remote across Europe or hybrid in Italy, on TypeScript, Next.js and React products.

Do you write code with AI?

Yes, and I review every line of it. Agents type, tests decide. Nothing reaches a merge unless I can explain it and a check can defend it.

Where are you based?

In Italy, working remotely with teams and founders worldwide, in Italian or English.

How do we start?

Send an email describing what should happen, what happens instead and which flow is blocking the release. You get an answer on whether it fits before anything starts.

~$ ./contact

Tell me about the project. Or the problem to solve.

Hiring, one broken flow or a product you want built properly: a short message with the context is enough to get a straight answer.