How it works
From one call to automation you own.
Four steps. Stop after any of them and keep what you’ve got.
The four steps
- 01
Talk
A free 30‑minute call. Show us the work on a shared screen.
Then a short written note, and an honest answer.
- 02
Scope
We watch the real work and prove the hardest step works.
Then a process map, acceptance criteria, and a fixed quote.
- 03
Build
We build it and test it on your real data.
It goes live when the results match your team’s. Most take a few weeks.
- 04
Run
We watch every run and fix it when a system or AI model changes.
One monthly fee covers hosting, monitoring, and fixes.
How we build it
- CodePlain, tested code for every predictable step.
- AIOnly where judgment is needed, tested on your own examples.
- A personAnything uncertain goes to your team, with the reason.
What you get at handover
- 01The code, in a repository you own
- 02Monitoring, alerts, and a review queue
- 03A runbook for running it without us
- 04A list of every system and login it uses
When something breaks
We’re alerted, we pause anything risky, and we fix it.
Then we tell you what happened and what we changed.
FAQ
Questions, answered.
01When are you not the right choice?
When a tool you already have can do it, or the volume is tiny. We’ll tell you on the call.
02What do you need from us?
Someone who knows the process, real past examples, and limited‑access logins to the systems.
03Do we own the code?
Yes. Once the build is paid, the code is yours, and it runs without us.
04What if it doesn’t work?
It runs, or you don’t pay for the build. If we miss the agreed criteria, you choose: more time or a full build refund.
05Can we take it in‑house later?
Yes, on 30 days’ notice, with no exit fees. We hand over everything and help your team.
Next step
Start with one process.
A free 30‑minute call with the engineer who’d build it.