Jacek Różański

Founder, The AI Mechanic

Senior backend engineer · Based in Katowice, Poland (CET) · Working with founders globally

// the short version

Who I am

I’m a senior backend engineer with 20+ years building production software — the kind that handles real users, real data, and real money without falling over.

I’ve worked across education, real estate, healthcare, government, automotive, sports, and adult. On each engagement I own the backend code, the infrastructure, the deployments, and the on-call.

I started The AI Mechanic because I kept getting the same email: “Our Lovable / Bolt / Cursor app is breaking in production. Can you take a look?” Enough times that it’s now a specialization, not a side project.

// what i do

Three things, in this order

  • Audit AI-generated codebases (Lovable, Bolt, Cursor, Claude Code, v0) and no-code projects (Bubble, Webflow) against production criteria — security, scalability, maintainability. The deliverable is a written document, sorted by business risk, not by code aesthetics. It’s yours to keep whether you hire me for the fix or not.
  • Fix what the audit finds — authentication holes, broken CORS, slow databases, silent data loss, sketchy third-party integrations. Real engineering work, not more prompts.
  • Ship the last 20% AI tools skip — domain setup, CI/CD pipelines, monitoring, backups, payment webhooks, email deliverability, disaster recovery. The boring-but-load-bearing stuff that separates a demo from a product.
// when to talk to me

What founders hire me for

  • Their MVP has started breaking in front of paying customers and they don’t know why.
  • A funding round or enterprise-customer conversation is approaching and the security questionnaire is blocking.
  • A non-technical co-founder shipped the product, and now an investor, board member, or first technical hire can’t evaluate what they actually have.
  • Their contractor left, and the codebase is opaque to everyone still on the team.
  • They’ve outgrown Bubble / Webflow / Supabase Edge Functions and need to move off without breaking their product.
// how i work

Three principles

Audit first

I don’t write code on day one. I read the code you already have and write you a document you can share with your team, your investor, or your next contractor. That document stays yours whether you decide to work with me afterwards or not.

Sort by business risk

Data-leak holes rank above aesthetic refactoring, every time. Compliance blockers rank above nice-to-have features. I’ll tell you what’s on fire, what can wait a quarter, and what’s fine to leave as-is for now.

No magic

Every finding is explained in plain English, in a way a non-technical founder can follow. If I can’t explain it to you without jargon, I don’t get to recommend it.

// the specifics

What I bring to the table

The short answer for founders or their technical advisors evaluating fit.

Technologies

  • Backend — PHP (Symfony, Doctrine), Node.js / TypeScript, Python. Domain-Driven Design, CQRS, event sourcing. REST, SOAP, OpenAPI.
  • Databases — PostgreSQL, MySQL, Aurora, Redis, DynamoDB, Sphinx.
  • AWS — ECS, EC2, Lambda, RDS, SQS, SNS, SES, S3, CloudFront, API Gateway, Route 53, ELB.
  • Frontend — React (Context API, React Query, React Router), TypeScript.
  • Architecture patterns — modular monolith over distributed microservices when the product doesn’t need them yet. Event-driven where it earns its keep.

Markets

  • AI-built MVPs — Lovable, Bolt, Cursor, Claude Code, v0 codebases taken to production.
  • No-code migrations — Bubble, Webflow, Airtable escaped to real infrastructure without breaking the product.
  • Education — corporate and public-sector training platforms.
  • Real estate — property-sales platforms, broker and buyer portals.
  • Healthcare — production systems in regulated environments.
  • Government — grant management, bulk-mail tooling, CMS for public agencies.
  • Automotive — dealer-facing and aftermarket SaaS.
  • Sports — B2B SaaS for the sports industry.
  • Adult — platforms with compliance, age-gating, and payment hardening.

Languages

  • Polish — native.
  • English — professional working proficiency. Discovery calls, engagements, and written deliverables in either.

Tools

  • Containers & infra — Docker, Terraform, Linux, Nginx.
  • CI/CD — Bitbucket Pipelines, GitHub Actions.
  • Observability — Sentry, Blackfire, Cloudflare, CloudWatch.
  • Testing — PHPUnit, integration-first over unit-first.
  • Source control — Git (GitHub, Bitbucket).
  • AI coding assistants — Claude Code, Cursor, Lovable, Bolt, v0. I use them; I also audit what they produce.

Education

  • Postgraduate Software Engineering — AGH University of Science and Technology, Kraków.
  • Master’s in Economic Computer Science — University of Economics in Katowice.
// next step

Want a second pair of eyes on your AI-built app?

The first conversation is free. I'll tell you honestly whether I'm the right help for where you are — and if I'm not, I'll usually have a suggestion for who is.

Book a free call →
Free consultation No obligation Reply within 24h