aceArt · AI Agency

AI consulting for companies

Most companies do not come unstuck because they know too little about AI. They come unstuck because they see too many possibilities at once and have no order of work. Our consulting sorts that out. At the end there is no trend report but a list with a number one, an honest effort estimate and a note on what data you need for it.

Get expert advice

The wrong first question

"Where can we use AI?" is the worst starting question there is. It leads to a collection of ideas that all sound plausible and none of which ever gets built. We ask it the other way round: where does somebody at your place regularly type data from one screen into another? Where does a customer wait longer for an answer than they should? Where do mistakes always happen at the same point?

Those places have an advantage: the effort, the source of error and the responsibility are already known there. You do not have to talk the benefit into existence, you can work it out. And it soon turns out that some of these cases need no AI at all but an interface. We say that too, even though it makes our own offer smaller.

What the consulting covers

Potential analysis

We look at your processes and search for the places with the best ratio of effort to effect. For that we talk to the people who do the process every day, not only to the management level. In our experience the people at the desk know the most expensive weak points better than any process diagram does.

Prioritisation

The collection becomes an order of work. Assessed against four questions: how much time does it save, how good is the data situation, how risky is a mistake, and how quickly is a first version visible? Plans that show something only after nine months are rarely the right ones to begin with.

Data situation and data protection

Which system is the source for which piece of information? That question decides between success and frustration long before a model comes into play. In parallel we settle which data may leave the house and which may not, and what that means for the choice of tools.

Roadmap

A plan that fits your people, not our calendar. With the things you can do yourselves, the things you need support for, and the things you should safely leave alone.

Consulting from people who also build

The difference between a recommendation and a recommendation you can act on is bigger than it sounds. Anyone who has never built an interface underestimates what it means to get data out of an ERP that has grown over years. Anyone who has never run an application does not plan for the running costs, which work differently with AI than with classic software: every request to a model is billed, so the bill grows with the usage.

We come from software development. That is why you get recommendations with a realistic effort figure beside them, and why we also say of some wishes that they will cost more in your system than the benefit justifies. If you then do the implementation yourselves or with another partner, that is fine with us.

From concept to trying it out

Consulting that ends with a document has a well-known fate. That is why we offer the next step along with it: in the MVP Workshop we build a clickable prototype of the prioritised use case together with you. In a single day a recommendation turns into something you can show around the house.

After that the route carries on, depending on the plan, to chatbots on your own company knowledge, to process automation or to integration into existing software. If your team first needs to feel sure with the tools, you are better off starting with training.

Funding for AI consulting

There are programmes at federal and state level for consulting and digitalisation projects, and AI consulting regularly falls under them. What counts is less the amount than the fit: some programmes are aimed at small companies, others assume a particular project description, and with some the administrative effort bears no good relation to the result.

In the first conversation we look together at which programme comes into question for your plan and whether the application is worth it. The application itself runs through the respective authority. We support the technical description, because the project description is usually the part that gets stuck. More on our funding advice page.

Frequently asked questions about AI consulting

What does AI consulting cost?
That depends on the scope: a potential analysis for one defined area is done a lot faster than a survey across several departments. In the free first conversation we cut the scope down and you get a written effort estimate before anything is commissioned.
What funding is there for AI consulting?
At federal and state level there are programmes for consulting and digitalisation that AI projects regularly fall under. Which one fits depends on company size, location and the kind of project. We check that with you in the first conversation and also say when the effort outweighs the benefit. The application itself runs through the respective authority.
Which processes are worth doing first?
The ones where somebody does the same work by hand regularly today and where a mistake is annoying but not dangerous. Classic cases are the pre-sorting of enquiries, reading out recurring documents and internal search across scattered knowledge. Anything to do with money, law or personnel decisions belongs at the end of the list, not at the start.
We have no data strategy at all yet. Is consulting premature then?
On the contrary, that is the best moment. The question of which system is the source for which piece of information has to be answered anyway. Whoever has answered it benefits from that even when no AI is used in the end.
Do you also advise without doing the implementation?
Yes. We say honestly what we recommend even when the implementation is cheaper in your own house. Consulting that only aims at the follow-up contract shows, and it is worth nothing.
How long does a potential analysis take?
For one defined area a few weeks is realistic, including the conversations with your people. The time needed sits less often with us than with finding dates in the house. If it has to be quick, we start with one area instead of all of them.
What is the difference from a classic management consultancy?
We do not only assess, we could also build it. That changes the recommendations: instead of a target picture you get a list with effort figures, and instead of a market overview the statement of what is feasible in your particular system.

Related topics

MVP Workshop

The fastest route from the recommendation to something clickable. To the MVP Workshop.

AI training

For when the team first needs to feel sure with the tools. To the training courses.

Funding advice

Which programmes come into question for digitalisation projects. To the funding advice.

Ready for your KI project?

Tell us about your project. We'll get back to you promptly with an honest assessment.

Contact us
HOME › AI-CONSULTING