What you get
- The fix, as a pull request against your repo — reviewable, revertable, yours.
- A test that proves it — fails on the broken code, passes on the fix. Runs on your CI, not just my machine.
- The evidence — before/after output, the diff, and the deployment or verification run.
- A written handover — root cause in plain terms, what changed, and anything I found but deliberately did not touch.
What it is not
- Not a rewrite, a redesign, or a migration.
- Not "general cleanup while I'm in there" — scope creep is what makes 48 hours a lie.
- Not new features. One blocker, named in writing before I start.
- Not regulated advice, and not work on a system I can't run myself.
How it runs
| Step | What happens |
|---|---|
| Before | You describe the blocker. I tell you honestly whether it's a fit — and say so if it isn't. We write the one-sentence scope together. |
| Hour 0 | Deposit clears, you grant access you control. The 48-hour clock starts. |
| Diagnose | I reproduce the failure and confirm the root cause before changing anything. If the cause isn't what we assumed, you hear it immediately. |
| Fix | The smallest change that resolves it, on a branch, with the test that proves it. |
| Handover | Balance clears, code and access transfer, evidence pack delivered. |
Is it a fit?
Good fit: an endpoint returning 500s in production, a webhook firing twice and corrupting orders, a build that won't deploy, a scheduled job that silently stopped, a migration that broke under real data.
Bad fit: "the app is generally slow and buggy", a greenfield build, or anything where nobody can reproduce the problem. Ask me — I'd rather turn it down than miss the deadline.
Not ready to commit to the full fix?
Emergency Audit — one working day
You get the root cause and a written fix plan you could hand to any engineer. If the work proceeds, this fee is credited in full against the engagement.
Who does the work
Luke Czak, senior engineer, working through Appaya Ltd. I use an automated build-and-test pipeline of my own to move fast — but the diagnosis, the judgement and the accountability are mine. You're buying a fixed outcome from a named person, not machine output.
Who is doing the work
Luke Czak. Thirteen years in product and engineering — Global Head of Technical Product Management at HSBC, product leadership at Zopa (FCA-regulated, 850k customers) and Smarkets, Head of Technical Product at Shieldpay (a regulated payments platform clearing £5–8bn a year), and CPTO at a startup taken from nothing to public launch in nine months.
I started as an engineer and never stopped. I read production code daily, and the reason I can fix a blocked release in 48 hours is that I have spent a decade in the kind of regulated systems where shipping the wrong fix is expensive.
I don't have client testimonials for this offer yet — it's new, and I'm not going to invent any. What I will do is walk you through my reasoning live before you pay anything, and the test I write for your fix is one you run yourself.
Terms, plainly
- A fixed fee, agreed in writing before work starts, plus VAT at the prevailing UK rate. 50% before work, 50% before code and deployment access transfer.
- A business outside the UK is invoiced without UK VAT under the reverse charge.
- GBP bank transfer (Faster Payments). Card is not offered for this engagement.
- You create the access; you revoke it when we're done. I don't take custody of your production credentials.
- Nothing is merged to a protected branch or deployed to production unless you ask in writing.
- IP in the deliverables transfers to you on cleared final payment.
- If I conclude it can't be fixed within scope, you get the diagnosis, the evidence and the fix plan — and owe no balance.
- 14-day warranty: if the same blocker reproduces, I fix it free.
- I publish nothing about your project without your written permission.
Start
Send the failing behaviour, what should happen instead, and how to reproduce it. You'll get a straight answer on fit — and a scope line to approve — before any invoice exists.