Challenge · New & unique

We want to use AI but do not know where to start

AI can take repeated enquiries and help on your own documents, but where does it make sense?

Sound familiar?

  • You know AI can do something, but not where to start
  • Experiments run into the sand without a clear problem
  • Repeated enquiries could perhaps be handled more cleverly
Employee looking at a screen with a chat and an orange AI symbol, considering where AI could start

What this is about

Everyone asks what you are doing about AI

Using AI means letting a model take a step that today needs a person to read, sort or phrase something. The board asks, customers ask, and there is no single right place to begin. The hard part is rarely the model, but choosing the one task where the gain can be measured.

  1. Why pilots stall

    Pilots rarely stall on the model. The ones we have seen stall did so because the data was not ready, because the pilot never made it into a real workflow, and because nobody had agreed in advance what success would look like.

  2. What the waiting costs

    You tie up budget and time in broad projects that never reach anything people notice in their working day. Meanwhile the same enquiries are answered by hand, week after week.

  3. When the answer is something else

    Messy data has to be tidied up before a model can use it, and that is not a pilot. A task with fixed rules is handled by ordinary automation, both more cheaply and more reliably than by a model.

A pilot is neither an AI strategy nor a system you have to adopt. It is one question answered on your own data, and an honest answer if AI is not the right tool here.

Only build what nobody sells

What makes you hard to copy, we build. The rest we have built before.

The safe route

How we start, without you committing to anything big

  1. Step 1

    Clarification meeting

    A short conversation about where there is most to gain, and what it is worth. Free and without obligation.

  2. Step 2

    Digital pre-project

    We find one small, narrow place in your own data where AI can help concretely. You get a concrete recommendation with numbers behind it, and if it does not pay off, we say so. A fixed price that is offset against the first stage if you build on within 90 days.

  3. Step 3

    Development in stages

    We build in stages at a fixed price and take whatever pays for itself fastest first. After that, operation and support from the people who know the solution.

You can stop after each step without the effort being wasted.

What the solution usually consists of

We decide exactly what is needed together

Exactly what it takes, we work out together in the pre-project. A challenge like this one we usually solve with one or more of these approaches.

  • Digital advice Strategic use of technology and support for your processes, so you get overview, control and better insight into your own data. It is part of most projects.
  • AI solutions Letting AI take repeated enquiries or find answers in your own documents, proven on a small scale first.

Ready to talk to us?

Contact us

A concrete example

What we would do here

A pilot on your data. We do not have a public AI case to point to, and we say so honestly. That is why we start with one small pilot on your own data and prove the gain before you build widely.

Frequently asked questions

We work with many different challenges, and they hang together more often than you would think.

See all challenges

Do you want to find your first AI step?

A meeting without obligation, where we look at your concrete situation. You get a suggestion for the next step and a clear sense of whether it pays off. You talk directly to the people who build the solution.
  • A reply from one of us by the next business day.
  • A no-obligation review of your situation.
  • Clarity on which options are relevant to you.
  • A concrete plan, if you choose to go ahead.

Contact us

We get back to you by the next business day.
Does this look like your situation?

Help us make this page clearer

Click the part that did not make sense — or where something was missing.

You are pointing at

Thank you. We have it.

Open it again if you spot something else.