AI consulting in Vienna. From someone who builds it himself.
You are wondering whether AI can take over a task in your business. I test that with your own examples and tell you plainly whether it pays off.
THE CASES PEOPLE BRING
Four tasks a check can start from.
Reading documents
Invoices or delivery notes arrive as PDF or photo, and someone types them in. The data should go straight into your system instead.
PDF or photobecomesData in your system
Sorting what comes in
Inquiries, tickets or orders come in every day. An AI should pre-sort and score them, while a person still makes the decision.
InboxbecomesSorted and scored
Preparing drafts
Many replies, quotes or reports look much alike. An AI should write the first draft before you check it and send it.
PatternbecomesDraft for review
Searching your documents
The answer is somewhere in your documents, but nobody finds it fast enough. An AI should answer and show the passage the answer comes from.
QuestionbecomesAnswer with its source
WHAT THE ASSESSMENT CONTAINS
At the end of the AI feasibility check you get a document with six chapters. Here is what each of them covers.
Task and data
Can a model solve the task? What data does it need, and do you have it?
Errors and sign-off
Which errors does the software catch, and which does a person see before sign-off?
Vendors compared
I test several vendors on examples from your business and put quality and cost side by side.
Sample values, unrelated to real vendors.
Cost at your volume
What a single item costs, measured on real examples and scaled to your volume.
Outages and operations
What happens when a vendor goes down or gets more expensive, and whether running the model yourself is worth it.
Recommendation
My recommendation and the reasons for it. If it is a yes, it also says how a prototype will be measured.
The outline is the same for every check, the content depends on your case.
HOW AI CONSULTING WORKS
It takes two steps to get to the answer. What happens next depends on how it turns out.
First call
You tell me what it is about and I ask questions. If the call already shows that AI will not pay off, I tell you right then and you save yourself the check.
AI feasibility check
You give me real examples from your business. I use them to check your case, compare vendors and write the assessment.
If it pays off
Prototype
Small enough to throw away and real enough to show how it handles your examples. How it will be measured is agreed before I build it.
Build and run
If the prototype does what it should, I turn it into the finished solution and take care of running it. More on the AI automation in Vienna page.
If it does not pay off
Then the work ends here
That is a result too. The assessment says why AI does not pay off in your case and whether there is a simpler way without AI.
ONE PERSON WHO ADVISES AND BUILDS
I only recommend what I can also build. You talk to the person who writes the assessment and builds the prototype. There is no handover between consulting and development, so nothing gets lost along the way. I offer this to small and medium-sized businesses in Vienna and across Austria.
The assessment is yours. You can also hand it to someone else and have it built there.
WHAT I CAN SHOW
Projects I have already built.
My product
A product of my own that reads receipts
Invoices arrive by email or as a photo. First pattern matching, then a language model with a fixed schema. A person confirms every suggestion, and the hit rate is measured.
My agent
An agent of my own that runs every weekday
It fetches inquiries from a source, scores them against a profile and files reports and drafts, which a person then sends.
My decision
A decision weighed in writing
For a product of my own I had to decide whether to run models myself or use a hosted service. I compared effort, licenses and cost per model.
Open source
Public tools
Some of what I build is public on GitHub: postkasten (opens a new tab), django-upgrade-report (opens a new tab) and a fork of the Keyword Planner MCP (opens a new tab).
QUESTIONS BEFORE THE FIRST CALL
What does AI consulting cost?
The first call is free. The AI feasibility check starts at €1,900 and takes one week. The exact price depends mainly on how clear the task is and how much data and how many systems are involved. A prototype gets its own quote.
Do you advise on the AI Act and data protection?
No, legal questions about the AI Act or data protection belong with a lawyer. What you get from me is the basis for that talk: which data goes to which vendor.
Do you run AI workshops or write strategy papers?
No, I do neither. I always work on one concrete case, up to the prototype. For training you are better off with trainers who do that full-time.
Are you an AI agency in Vienna?
No, I am a developer and consultant from Vienna. There is no sales team and no project management, you talk to me directly.
Tell me about your case.
After the call you will know whether a check is worth it.
- First call
- free
- Check
- from €1,900
- Result
- in one week