SERVICE 06 — OPERATIONS & SUPPORT

So that what was built once still runs a year from now.

Automations, agents, software and websites need looking after: tools change their interfaces, models get retired, requirements grow. I take on monitoring, updates and further development, and I'm the person who picks up when something jams.

An arch of black stones whose blue keystone at the top holds the whole structure together.
Same day response on working daysMonthly cancellationOne point of contactFixed price per month
OPERATIONS & SUPPORTMONITORINGMAINTENANCEUPDATESBACKUPSFURTHER DEVELOPMENTONE PERSON TO CALL
EXAMPLE: THE SILENT FAILURE

A provider changes its interface, the automation keeps running but stops transferring data. Without monitoring, something like that only surfaces when someone misses a number, often weeks later. With monitoring the alert arrives within minutes, not through a customer call.

WHEN SUPPORT PAYS OFF

Automated processes don't fail loudly, they fail quietly.

When a person forgets a task, somebody notices. When an automation fails, nothing happens at first: no alert, no error on anyone's desk. That's exactly why everything running in the background needs someone who looks before the customer does.

A

Processes run in your company that nobody looks at daily.

B

Nobody on the team feels responsible for updates and error messages.

C

There's no one to call when something stops working.

D

The setup should keep growing instead of staying the way it was built.

What support actually covers.

WHAT'S INCLUDED
MONITORING

See errors before you notice them

Every process reports in when it doesn't complete. I look at the alert and get in touch if you need to do something.

UPDATES

Keep systems current

Security updates, version changes and changed interfaces, before they become a problem day to day.

CHANGES

Small changes on request

New recipient, changed wording, one extra step: small things are covered by the support plan.

GROWTH

Add new processes

When another process comes along, it's planned as a stage and delivered at a fixed price.

AI UPKEEP

Sharpen agents and models

Keep the knowledge base current, improve answers based on real cases, switch models when better ones land.

SECURITY

Backups and access

Regular backups, controlled permissions and an orderly process when someone leaves the company.

You have something someone else built? I'll take that on too, after a short inventory. Tell me briefly what's running and you'll get an honest read.

What a silent failure costs.

THE NUMBERS
SAMPLE CALCULATION: FAILURE WITHOUT MONITORING
Affected cases per week35
Went unnoticed for3 weeks
Manual rework12 h
Cost of rework (assuming €60)€720
Support plan€190/month
One incident equals4 months

Not counted: the enquiries that went unanswered in that time, and the effort of working out afterwards which cases were affected at all.

The same incident, with and without support

STEPNO SUPPORTWITH SUPPORT
NoticingWhen somebody misses somethingAutomatic alert, immediately
UnderstandingWho did this, what's broken?The cause is in the log
FixingFind a provider and brief themSame working day, no briefing
ReworkCatch everything up by handCatch-up run triggered automatically
AfterwardsCan happen againA check is added so it surfaces
TotalDays to weeksUsually the same day

How a support plan starts.

HOW IT RUNS
1

Inventory

What runs, where does it run, who has access? Even when someone else built it, I need a complete picture first.

2

Secure it

Set up monitoring, alerts and backups, remove the obvious risks. That's the part with the fastest effect.

3

Ongoing operation

Updates, checking alerts and small changes. You write to me directly, with no ticket system in between.

4

Keep developing

Once a quarter we look at the numbers and open wishes together and decide what comes next.

TECHNOLOGY

I look after everything I'd also build.

Automations, agents, custom software, websites and shops. Whether I built it doesn't matter, as long as I get access and the setup is comprehensible. After the inventory I'll tell you what I can take on and what should be cleaned up first.

n8n · Make · ZapierAI agentsOpenAI & ClaudeNext.js & ReactWordPress & WooCommercePostgreSQLHosting in the EUBackupsMonitoring & alertsAccess management

How you can tell it's working

IN PRACTICE
Extremely happy! We finally look professional when bidding for contracts.
Vasil StoyanovSPEDO GMBH
FROM PRACTICE

Most incidents never reach the customer

The normal case in support isn't the emergency, it's the alert that came in during the morning and is dealt with by lunchtime. It only becomes visible when you compare how often someone previously had to fix something themselves.

Talk through what you have →
SAME DAY

Response on working days, usually much faster.

NO TICKET SYSTEM

You write directly to the person who fixes it.

YOURS TO KEEP

All accounts and data stay in your name.

CANCEL MONTHLY

No minimum term, no notice periods.

What does support cost?
A fixed price per month.

ALL PRICES EXCL. VAT

Basic

from €190/month

Cover for individual processes or a website.

  • Monitoring and alerts
  • Security updates and backups
  • Troubleshooting on working days
  • Small changes within scope
  • Cancel monthly
Request Basic
MOST OFTEN CHOSEN

Operations

from €390/month

For everything that runs in the background and has to keep running.

  • Everything in Basic
  • Several automations and systems
  • Adjustments when tools update
  • Agents and knowledge base maintained
  • Quarterly session on what's next
Request Operations

Partner

from €790/month

When digitalisation is meant to keep going.

  • Everything in Operations
  • Fixed development time each month
  • New processes delivered continuously
  • Priority on incidents
  • Reachable via chat and phone
Request Partner

The price follows how many processes are covered and how critical they are. Larger stages are billed separately at a fixed price, so the monthly support stays predictable.

HONESTLY

When you don't need a support plan.

Support should pay off, not just tick along. In these four cases I'll tell you to save the money.

01

It's just a simple website

For a site with no shop and no integrations, hosting with automatic updates and backups is enough. You don't need a contract with me for that.

02

Someone on the team does it

If someone at your company looks after technology and has time for it, I'd rather hand over cleanly, documentation and briefing included.

03

The process isn't critical

If a two-day outage bothers nobody, calling when something breaks is enough. I'll bill that individually.

04

Nothing has been built yet

Support only makes sense once something is running. Before that it's about consulting or building, not a monthly contract.

BACKGROUND

Running and maintaining automations, agents and software

By Chris Pump, developer since 2021Last updated: August 2026

Most digitalisation projects get planned, built and then left to themselves. That works for a while, because software rarely fails overnight. It frays: a changed interface here, a retired model there, an access token that no longer applies. A year later only part of it still runs, and nobody knows exactly which part.

Why silent failures are the real risk

A server that's unreachable gets noticed immediately. The dangerous one is the process that keeps running and just skips a step: invoices without an attachment, enquiries that no longer reach the CRM, reports with last week's numbers. Those errors often stay unnoticed for weeks and cause more rework than a clean outage. That's why every process I build reports success, not just failure: if the signal stops, that itself is the alert.

What maintenance means for AI systems

With classic software, maintenance mostly means updates. With anything built on language models, something else comes on top: providers retire models, prices change, and answers shift with every version. Then there's the content side, because a knowledge base with last year's prices is worse than none. A sensible rhythm is to spot-check real cases regularly and update the knowledge base whenever something in the business has changed.

Access, dependency and handover

Support must not create dependency. So accounts, servers and repositories run in your name, and I work with my own credentials that you can revoke at any time. Every support plan includes documentation that's kept current, so someone else can take over if needed. That makes cancelling an appointment rather than a project: revoke access, the docs are there, on you go without me.

Why Kasoria

At Kasoria you write to the person who knows the system, with no ticket system and no support tiers. I also look after things I didn't build myself, after an inventory, and I'll tell you honestly when something needs cleaning up before it can sensibly be supported.

FAQ

Common questions about operations and support.

How fast do you respond to an incident?

On working days the same day, usually within a few hours. Critical outages take priority over whatever else is on my desk.

Do you support things someone else built?

Yes, after an inventory. I look at how it's put together and what state it's in. Then I'll tell you what I can take on and what should be cleaned up first.

What counts as a small change?

Anything done in roughly half an hour: wording, recipients, thresholds, one extra step. Larger extensions get planned as a stage with their own fixed price.

Is there a minimum term?

No. Support can be cancelled monthly, with no notice period and no exit conversation. On cancellation you get documentation and access handed over in order.

Who has access to my systems?

Only me, with my own users and the permissions needed for the job. Accounts and servers run in your name, and you can revoke access at any time.

What happens when you're on holiday?

I announce time off in advance. For critical systems I set up cover or a manual fallback beforehand, so nothing depends on my availability alone.

Are hosting and licences included?

No, those run through you directly so you stay independent. I set everything up and keep an eye on the costs, but the contracts sit with you.

Do I get a report?

On request, monthly and in short form: what failed, what changed, what's coming up. Plus a quarterly session where we talk about what to build next.

Can I switch tiers or pause later?

Yes, you can change between tiers at any time. Pausing is possible: monitoring keeps running, and changes are billed individually during that period.

What about emergencies at the weekend?

For systems where a weekend outage genuinely hurts, we agree that separately upfront. The standard is working days, because anything else honestly costed means money most companies don't need to spend.

Do you bill hours that don't occur?

No. Support is a fixed price for availability, monitoring and ongoing upkeep, not an hour budget that expires. Larger projects are commissioned separately.

What if I end the support plan?

You get the documentation and all access, and the systems keep running unchanged. On request I hand over directly to your new provider or to your team.

WHAT HAPPENS NEXT?

Tell me what's running at your company and who looks after it today.

30 minutes, free and without obligation. Afterwards you'll know what needs covering, what that costs and what you can skip.

Usually answered the same working day.