Process automation that works in everyday use.
I automate business processes in small and mid-sized businesses that are still done by hand today: from the customer enquiry to reporting. First we clarify whether and where automation pays off. Then I implement the solution that actually makes the process simpler.
Reply within 24 hours. Free of charge and without obligation.
- Simplifying, digitalising and automating recurring business processes
- Enquiries, documents, quotes and orders, data transfer, approvals, reporting
- Automation, connected systems or AI, depending on what the process needs
- Your existing software is used and connected, not replaced
- A process that measurably runs more simply, with a fixed end to the project
- Specific process: straight into implementation. Unclear what pays off: the Process Check first
01 Starting point
Many manual steps. One process.
Manual processes rarely cost a lot of time in one place, but a little in many places. Automation turns them into one continuous process.
02 Processes
Where automation comes in at the office.
Six processes in which many companies lose time every day. The routine can run automatically, the decisions stay with your team.
| Process | Today | Automated | Stays with people |
|---|---|---|---|
| Customer enquiries | Enquiries arrive by email, form or phone. Someone reads and sorts them, transfers the details into the CRM and asks for whatever is missing. | Enquiries are recognised, the details are captured in a structured way and assigned. Missing information is requested specifically. | Assessment, priority and the reply to the customer. |
| Documents and mail | Orders, invoices, forms or letters arrive as PDFs or on paper. The data is typed in, deadlines are noted down, the file is filed by hand. | The data is read from the document, checked and saved in the right system. Deadlines are recognised, and the document is filed where it belongs. | Discrepancies and documents that are not clear-cut. |
| Quotes and orders | Details from the enquiry, price lists and earlier quotes are gathered together. Order confirmations are written by hand. | A draft quote is created from the existing data. Once accepted, the order is created; open quotes are put forward for follow-up. | Costing, price and approval. |
| Data between systems | Whatever is in one system is transferred into the next: customer data, orders, hours, receipts. | The information is captured once and passed on automatically, with a check for completeness. | Clarification when data does not match. |
| Internal approvals | Purchase orders, incoming invoices or expenses wait in an inbox for approval. Nobody knows exactly where the case is. | The case goes to the responsible person with all documents. A reminder follows if it is delayed, and the status is visible at all times. | The decision itself. |
| Deadlines and reporting | Deadlines, follow-ups and reminders are maintained in lists by hand. Every month, figures are compiled in Excel from several sources. | Deadlines and follow-ups reach the responsible person in good time. Reports are generated from the existing data and arrive on a fixed date. | Analysis and consequences. |
These are examples, not packages. Which process matters in your business is something we work out case by case.
In practice: customer enquiries via messenger03 Potential
You can spot automation potential in everyday work.
It rarely shows at first glance, but in small things that come back every day.
Signs
- Someone types in what is already available digitally.
- The same information sits in two systems, and not always identically.
- Cases wait because someone has to look something up, ask or approve.
- Mistakes only come to light when the customer calls.
- At the end of the month, figures are gathered rather than analysed.
- More orders mean almost as much more work in the office.
Questions
- Does the process occur often enough?
- Does it follow rules that can be written down?
- Is the information available digitally?
- How many exceptions are there, and who decides them?
- Will the process stay stable for a while?
In depth under Insights
04 Check first
Check first. Then automate.
Not every process should be automated. It pays off where effort, savings and implementation are in a sensible proportion.
-
Understand
How does the process really run today?
-
Spot the potential
Where do manual work and media breaks occur?
-
Assess the economics
What does it cost, what does it save, when does it pay off?
-
Automate
What pays off, in stages.
-
Measure the result
What has changed?
What is clarified first
- Which processes actually cause unnecessary effort?
- Where are there media breaks and recurring manual work?
- Which processes are technically suited to automation?
- How large is the potential saving, how complex the implementation?
- When does the investment pay for itself?
- Which process comes first?
The calculation behind it
- Time per case × cases per year × internal hourly rate
- Standard cases per year × minutes saved per standard case ÷ 60 = hours saved per year
- One-off effort ÷ (annual savings − annual running costs) = years to payback
In the Process Check, with values from your everyday work instead of estimates.
Fixed price from €750 plus VAT, results after about two weeks.
Already clear on the process and the goal? Discuss your project directly
05Approach
Not the biggest solution. The most sensible one.
I start with the smallest step that solves the problem.
Often that is not software yet, but a clearly organised process. And sometimes the right decision is not to automate a process at all. I will tell you that, too.
-
Simplify
Unnecessary steps are dropped, responsibilities become clear, templates consistent. That alone often gains a lot.
-
Use what you have
Your software can already handle the step if it is set up accordingly.
-
Automate by rules
Recurring steps run according to fixed rules: transfer, check, assign, remind, file.
-
Connect systems
Information flows between email, CRM, ERP and accounting instead of being typed in again.
More on system integration -
Use AI
Where content is written freely, in emails, documents or notes, AI picks out what matters. AI agents take on multi-step tasks, a knowledge system makes company knowledge easy to find.
More on AI agents -
A dedicated application
If no existing software fits: a lean web application, made for exactly this process.
Each step is a possible result. The next one is only added if the previous one is not enough.
06 Rules or AI
Rules where possible. AI where it helps.
Many processes don’t need AI. Some can only be automated sensibly with it.
| Rule-based automation | AI as a building block | |
|---|---|---|
| Fits when | the input is structured: form fields, tables, data from a system. | the input is written freely or inconsistent: emails, PDFs from different senders, notes. |
| How it works | Fixed rules: if this, then that. Every step is predictable. | Recognises content, assigns, summarises, suggests. The result is very likely to be right, but not guaranteed. |
| Example | New order in the online shop → order in the ERP → confirmation to the customer. | Enquiry by email → recognise product, quantity and requested date → prepare the case in the CRM. |
| Control | The rules are checked once and then always run the same way. | Checks and clear limits. Uncertain cases go to a person. |
| Running costs | Operating and maintaining the automation. | In addition, use of the AI service, depending on volume. Belongs in the calculation from the start. |
Often it is a combination: AI reads, rules decide, a person checks the exceptions.
AI agents go one step further: they complete multi-step tasks on their own, within defined limits. They make sense where a process has too many variants for fixed rules. An AI knowledge system gives them and your team the company knowledge they rely on.
07 Systems
Your systems stay. They get connected.
Workflow automation builds on what you use today. New software only if it actually solves something better.
-
Email and calendar
Outlook and Microsoft 365, Google Workspace, shared inboxes
-
Documents and filing
PDFs, scans, SharePoint, OneDrive, network drives
-
Spreadsheets
Excel lists and shared spreadsheets that serve as a database today
-
CRM
Customer data, enquiries, sales opportunities
-
ERP and inventory management
Items, orders, delivery notes, invoices
-
Accounting
Receipts and the steps leading up to the hand-over to DATEV, your accounting software or your tax adviser
-
Forms and website
Contact and enquiry forms, customer portals
-
Industry software
Specialist solutions with an interface or import and export
-
Teams and messengers
Notifications and approvals where your team already works
What it takes is reliable access: an interface (API), an import and export, or another stable route. I check whether it exists before anything is built, not afterwards.
More on system integration08 Project
From the first conversation to a process that runs.
You know which process should become simpler? Then we go straight into implementation, with a defined result and a clear end.
-
Initial call
30 minutes by phone or video, free of charge. We talk about one specific process and work out where it makes sense to start.
-
Understand the process
We look at how the process actually runs today: steps, systems, volumes, exceptions. If it is still open whether a change is worth it, a Process Check of your automation potential will clarify that. You can prepare with the Process Profile.
-
Define goal and scope
We define how the process should work afterwards, how we will measure success and what is explicitly not included. This results in a quote with a clear scope and a fixed end to the project.
-
Implement and test
I build the solution in stages and test it with real cases from your everyday work before it goes live.
-
Introduce
Your team gets to know the solution, responsibility for exceptions is clear, the documentation is in place.
-
Measure and hand over
We compare against the starting value and close the project with a hand-over.
09 Afterwards
A process that still runs a year from now.
Automations need looking after, because systems, interfaces and processes change.
- Documentation and hand-over
- You know what the solution does, where it runs and who is responsible for exceptions.
- Monitoring
- If something does not run as intended, there is an alert, instead of cases quietly getting stuck.
- Measuring
- After a few weeks we compare against the value from before. For the baseline there is the “Measure First” tracking sheet.
- Developing further
- If you wish, I continue to look after the solution: adjustments when something changes, and new steps when they pay off. That is Ongoing Process Optimisation, not a mandatory contract.
10 Questions
Frequently asked questions.
On when it pays off, costs, AI, existing software and what happens when something goes wrong.
What does process automation mean in practice?
Process automation means that recurring steps of a business process run without manual work, such as transferring data, reading documents, assigning cases, reminding of deadlines or creating reports. A person steps in where experience, responsibility or a decision is needed.
Workflow automation usually means the same thing with a focus on individual workflows. Process digitalisation is often the step before: information is available digitally rather than on paper or in individual people’s heads.
Which companies is process automation useful for?
For small and mid-sized businesses where enquiries, documents and data pass through the office every day: technical businesses, trades businesses with their own office organisation, B2B service providers and trading companies.
What matters is less the size of the company than the process itself: it should occur often enough, follow rules and be based on information that is available digitally.
Do we need to introduce new software for this?
Usually not. The starting point is your process, not a particular tool. Existing systems are used and connected, not replaced. New software or a small dedicated web application is only considered if no existing solution can handle the process sensibly.
When is process automation worth it?
When a process occurs frequently, mostly follows clear rules, is based on information that is available digitally and has few exceptions, and when the effort saved over its lifetime is greater than building, running and maintaining the solution.
If one of these conditions is missing, simplifying or making better use of existing software is often the cheaper answer.
Do we have to start with a Process Check?
No. If the problem and the goal are clear, we go straight into implementation. A Process Check makes sense when it is still open whether and where a change is worth it: which processes cause unnecessary effort, what automation would save, what it would cost and where to start. Where to start is something we clarify in the free initial call.
What does process automation cost?
That depends on the process, which is why there is no price list. Before anything is implemented, we define the scope, the goal and a fixed end to the project, and you receive a quote for it.
If it is still unclear whether and where a change is worth it, a Process Check will clarify that. It costs 750 to 1,250 euros plus VAT, depending on whether we look at one process or several.
Running costs are quantified before implementation: licences for the software used and the use of AI services, depending on volume. Support after the project is optional and starts at 250 euros a month. It is not a mandatory contract.
How long does an automation project take?
That depends on the scope. I deliberately size a first project so that it can go live within a manageable time and be measured. We set the schedule together with the scope before we start.
What is an AI agent, and do we need one?
An AI agent is software that uses a language model to plan and carry out several steps on its own, such as searching for information, operating programs and entering results into a system, within limits defined beforehand. Unlike rule-based automation, it does not follow a fixed, pre-programmed sequence.
Many processes don’t need that: fixed rules are simpler, cheaper and more predictable. An AI agent pays off where a process has too many variants for fixed rules.
What happens if the automation makes a mistake?
Every solution has checks built in. Unclear or incomplete cases go to a person instead of being processed incorrectly, and faults are reported. Who in your team is responsible for exceptions is something we define during the introduction.
How do you avoid automating too much or making it too complicated?
We define beforehand which problem is to be solved and how we will measure success. Then I start with the smallest step that solves the problem: simplify, use existing software, and only then automate. If automation doesn’t pay off, I say so, even if the recommendation is: build nothing.
How do you handle data protection?
Which data is processed, where it is stored and which services are involved is something we clarify before implementation. For AI services, this includes the provider, where the data is stored and a data processing agreement. Wherever possible, the data stays in the systems you already use.
Get in touch
Which process costs you the most time today?
Describe it to me in the initial call. Then we will know whether to implement straight away or to check first, with a Process Check, what pays off.
Reply within 24 hours. Free of charge and without obligation.