Rescue Engineering
From vibe-coded prototype to production software.
Your MVP got you users, maybe revenue. It was built fast, with Cursor or Lovable, by an agency or by a founder working nights. Now it's slow, it breaks in ways nobody can explain, and every change feels like it could take something else down.
- Signs you need this
- Deploys break things that worked yesterday.
- Nobody on the team fully understands the codebase.
- You're putting off hiring engineers or signing bigger customers because you're not sure the code will hold.
- You suspect API keys are sitting in the repo, or that one user could see another user's data.
- A customer or investor asks about security or data handling, and you don't have a good answer.
- What I do
- Assess. Go through the code, cloud setup, database schema, login and permissions, and how secrets and customer data are handled.
- Stabilize. Fix the riskiest problems first, like security holes, leaky data access and deploys that keep breaking. Then add a robust automated test suite, a CI/CD pipeline that runs it on every merge, and alerts that catch errors before your customers do.
- Hand off. Plain-English docs and runbooks for deploying, rolling back and handling outages, so your team, or your next hire, can take over.
- You get
- A written risk assessment that ranks each problem by severity and effort to fix, and says what can safely wait. After that, a codebase you can change without holding your breath, with tests and docs your next engineer can pick up.
- Typical shape
- A two-week assessment first. If you want to go further, a fixed-scope stabilization project built from what the assessment found.
- Why me
- I've launched 10+ production applications, built engineering teams from scratch at three companies and owned PCI DSS compliance for a payments platform. I know which shortcuts are fine to keep and which will hurt you, and I won't push a rewrite when targeted fixes will do.
Next step
You'll get a straight answer on what to fix first, whether or not we work together.