Managing technical partnership

The work continues after the first system is live.

After the first system goes live, kratt can stay on as your technical partner. We review the running system, choose the next justified priority, improve it and keep the decisions documented. When your team is ready to run it, we hand it over with the code, data, documentation and a named owner for the next decision.

kratt · technical partnership
Operating loop
  1. Reviewstart
  2. Next priority
  3. Improve
  4. Clear ownership
Logbook
  1. system cardAutomation spine: 21 workflows, 14 active
  2. case studyBackup loop: weekly restore test, Postgres 16.15
  3. case studyRegistry radar: daily diff against yesterday, 376,365 entities
Entries come from three of our case studies. The date is the system card’s, or the day the case study was last reviewed.
Four steps

How the technical partnership works

We can stay involved as the technical partner, keeping the system useful while your priorities, tools, and operating reality change.

  1. Review what is running.

    We look at the agreed measure, the system health, and where people still need to work around the process.

  2. Choose the next priority.

    The roadmap turns the evidence into one clear technical decision, not a list of disconnected ideas.

  3. Improve the system.

    We implement the next piece, connect it to what already exists, and keep the decisions documented.

  4. Keep ownership clear.

    We can continue to run the work, support your team, or hand over a documented system when that is the right next step.

Our own systems

Systems we run ourselves, every day

Three systems we built and run for our own work. Every number comes from that system’s case study, with the case study’s date next to it.

  • Backup and restore test

    0 production writes during restore testing

    Once a week a restore test loads the dumps into a throwaway Postgres and destroys it after proving the data loads. The version recorded for recovery is Postgres 16.15.

    Case study reviewed 2026-09-16

    Read the case study
  • The automation spine

    21 workflows on the 2026-08-17 system card, 14 of them active

    Our workflows run on one shared n8n instance, with Postgres state and a private admin path, instead of one-off scripts on laptops.

    Case study reviewed 2026-09-30

    Read the case study
  • Registry radar

    376,365 entities in the recorded extract

    Every day it diffs the Estonian Business Register open-data extract against yesterday’s and reports new registrations, new VAT numbers and status changes.

    Case study reviewed 2026-09-30

    Read the case study
Ownership

Documentation and handover

You decide which path fits. Either way, the decisions and who owns what are written down.

Path 1

We keep running it

We can continue to run the work and support your team.

As your partner we handle:

  • monitoring
  • improvements
  • roadmap decisions
  • documentation
Path 2

We hand it over

If your team takes the system over, we leave a documented operating path rather than a black box.

Written down:

  • code
  • data
  • accounts
  • documentation
  • maintenance
  • the person responsible for the next decision
FAQ

What people ask about the technical partnership.

What happens after launch?

We can stay involved as your technical partner, through monitoring, improvements, roadmap decisions and documentation. When your team is ready, we hand the system over.

Can our own team take the system over?

Yes. If you take it over, we leave a documented operating path rather than a black box. kratt can also remain the managing technical partner for monitoring, fixes and roadmap decisions.

Who owns the code, data and accounts?

You keep control of your accounts, data and permissions. At handover we name the important parts in writing: code, data, accounts, documentation, maintenance, and the person responsible for the next decision.

How do you handle backups?

We maintain weekly Git bundle backups for the brain and active repositories and have tested restoring from a backup. We do not present the current setup as automated off-site disaster recovery, a certified business-continuity programme, or a universal uptime promise.

The audit

Start with the audit. Then decide what is worth building.

Karl leads a 30-minute conversation about how the work runs today, where value is being lost, and what the next responsible step looks like. If a build makes sense, the scope follows the findings. If it does not, we will say so.

Book your free 30-min audit call
kratt

kratt takes AI in companies from an initial idea to a reliable system. Based on the audit, we define where AI pays off and start development where the return is highest.

Book the free audit call
The newsletter

Occasional notes on
what’s actually working.

No spam. Cancel anytime. Occasional notes only.
Book your free 30-min audit call