Process digitalisation for mid-sized companies

Digitalise your processes – one at a time

Most companies have workflows that cost time every single day: data is retyped, lists are reconciled by hand, enquiries are sorted out of an inbox. We take one of them and turn it into software that runs in day-to-day work. We use AI where it does the job better, and nowhere else.

How it works

Initial call free of charge. Process analysis for one process: 1,200 € net, fixed price.

Where it usually gets stuck

Rarely a lack of technology. Mostly work nobody ever looked at as a process, because it has always been done that way.

Retyping

What sits in one system gets entered again in the next. Every day, by hand.

Excel as a system

A spreadsheet only one person still understands carries an entire workflow.

Knowledge in the inbox

The answer already exists. It is in an email from two years ago.

Systems that don't talk

ERP, shop and accounting know nothing about each other.

What we build from it

Not a platform and not a kit, but a solution for exactly this workflow that fits into the systems you already have.

Integrations

Connections between the systems you already have, so data is entered only once.

Document processing

Read details from invoices, orders and forms, check them and pass them on to the right system.

Internal tools

A small application with roles, validation and history that replaces a spreadsheet that has grown over the years.

Search across your own documents

Ask questions of manuals, quotes and documentation and get an answer that points to where it was found.

Agents for routine tasks

Software that completes several steps on its own. Before anything leaves the building, a person approves it.

If you need a larger application with its own roles and data model, such as a customer portal, custom software development is the right starting point.

Where AI belongs and where it doesn't

If a workflow can be described with fixed rules, we write rules. That is cheaper, faster and gives the same result every time.

AI pays off where the input is messy: free-text emails, scanned documents, paperwork in twenty different layouts. There a language model can read, classify and summarise what would otherwise need a person.

Whether a model is good enough for your case is something we test with your own examples before we commit to anything. And where a mistake would be expensive, a person stays in the approval step.

Cloud, Europe or your own hardware

Where a model runs is a decision per process, not a matter of belief.

Cloud models

The most capable models, quick to set up, billed by usage. The data goes to the provider for processing.

European providers

Processing in data centres in the EU. The commitments differ by provider and contract, and we read through them with you.

Local models

Run on your hardware or on a rented server. That means fixed costs and operating effort, and not every task works equally well with them.

Local is not automatically the better choice. Which option fits depends on what data is processed, how much of it there is and how good the result has to be.

n8n or custom code

For manageable workflows we use n8n, a tool in which the individual steps are visibly wired together. The instance runs on your hardware or in your data centre and belongs to you.

When things get more complicated, we write custom code, with tests, version control and documentation. That is the craft we come from.

Either way: what we build belongs to you, and you can keep working with it without us.

From the first call to a running process

  1. 1

    Initial call

    30 minutes, free of charge. Afterwards both sides know whether there is a process worth a closer look.

  2. 2

    Process analysis

    We map one process together with the people who do it every day. You get a written result: how it runs today, what can be automated and what cannot, which model option fits and what a pilot is likely to cost. Fixed price: 1,200 € net.

  3. 3

    Pilot

    One process, with real data. Beforehand we agree how we will measure whether it holds up, and when we stop.

  4. 4

    Operation

    Handover with documentation. After that we look after the solution billed by effort, within our service hours, Monday to Friday, 8 am–6 pm.

When we advise against it

Not every workflow gets better by automating it. In these cases we say so:

  • The workflow only comes up a few times a year.
  • There is off-the-shelf software that already does it.
  • The data everything would build on is patchy or contradictory.
  • The decision depends on experience and judgement.

Then the process analysis says exactly that, and you have spent 1,200 € instead of a project budget.

Where we stand

We have been building software, APIs and integrations for years. You can find those projects under Projects.

We use AI and automation in our own work, with cloud models and with local ones.

We cannot show client projects in this area yet. That is why we start small: one process, a fixed price for the analysis, a pilot with a stop criterion.

What we do not promise

No savings figure: what a process delivers after the change is something we only know once it is measured. No legal assessment: we name what needs clarifying, the judgement belongs with your lawyer or data protection officer. And no funding promise: programmes exist, you submit the application yourself, before placing the order.

Frequently asked questions

What does a process analysis cost, and is it credited?

The process analysis costs 1,200 € net as a fixed price and covers one process. If you then commission the pilot from us, we credit the full amount. If you decide against it, you keep a written result you can take elsewhere.

What does a pilot cost?

A pilot is billed by effort at our hourly rate of 120 € net. The process analysis states how many hours we expect. That is an estimate, not a commitment. On top come running costs for the model or the server, which we name in advance.

Do we need AI for this at all?

Often not. Many workflows can be solved with fixed rules and a clean integration, which is cheaper and more reliable. AI pays off where the input is messy, such as free-text emails or scanned documents. Which of these applies to your process is part of the process analysis.

Does our data stay in-house?

That depends on the option. If the model runs on your own hardware, the data does not leave your network. With cloud models it goes to the provider for processing. In that case we clarify beforehand where processing takes place and what the contract says about how the data is used. This is not a legal assessment, which belongs with your lawyer or data protection officer.

Who owns the solution in the end?

You do. You get the source code and the documentation, and an n8n instance runs on your hardware or in your data centre from day one. You can develop the solution further with us, run it yourself or hand it to someone else.

Do you work with our existing software?

Yes, that is the normal case. We connect to what is there, through APIs, exports or the database. Whether your system offers a usable way in is something we clarify in the process analysis. If it does not, we say so before a pilot begins.

Is there public funding?

Germany's federal and state governments fund digitalisation projects, in Bavaria for example through the Digitalbonus programme. Whether your project qualifies is decided by the funding body, and we cannot promise anything. The order matters: you submit the application yourself, and you do so before placing the order. From us you get the quote and the scope of work for it.

Which workflow costs you the most time?

Tell us about it. In the initial call we work out whether a process analysis is worth it. Sometimes the answer is no.

Prefer to talk? +49 9163 6791505 (Mon–Fri, 8 am–6 pm CET)