Skip to content
Superintelligent
SVEN
Log in Book a meeting

Home / About us / The endgame

The endgame

We make ourselves dispensable. On purpose.

We have no interest in you needing us forever. Our job is to build your own AI capability, so that you can manage without us. That is how we earn your trust.

How we measure it ↓ · Where do you want to start? ↓

Sketch of the principle: our involvement goes down and your own AI capability grows, step by step from the start to after the handover. START TRAINING BUILDING HANDOVER AFTER
Your AI capability Our involvement A sketch of the principle, not measured data

The fair question

Why would a consultancy want to become dispensable?

We want to make money too. But not by locking anyone in. There is more AI work to do than we could get through in the rest of our lives, so we do not need to hold on to anyone to stay busy.

There is also a reason that has to do with where the value comes from. According to BCG, roughly 70 per cent of the value in AI initiatives comes from people and processes, 20 per cent from data and technology and only 10 per cent from the algorithms themselves (BCG, 2026).

So the capability has to sit with you, in your people and your ways of working. It cannot be rented. That is why we spend our time building it, and handing it over.

How we work · Thomas explains (in Swedish)

AI becomes part of everyday work, not a project on the side

01 Where are you losing time?

We map where time, knowledge and money disappear into manual steps and dependence on individuals.

02 What has the biggest effect?

We find the combinations of your expertise, your processes and AI that have the biggest effect.

03 What should you start with?

You get each opportunity rated: business value, feasibility, risk, ownership and time.

How we measure success

How do we know you can manage on your own?

A promise to make yourself dispensable is worth nothing if it cannot be measured. We measure on two tracks, and then the thing that matters most: what happens when we step back.

Track 1 · Efficiency AI

From hours to solved tasks

  • The AI interviews at the start show where the time goes, for example how many hours a week different tasks take.
  • Afterwards we follow up the same tasks. Are they solved with what employees learned in the training, with automations or with agents they built themselves?
  • We follow how much your AI licences are actually used in the platform you already have, and how that changes with what we do.

The hours are the employees’ own estimates. That is why we also follow actual use in the platform.

Track 2 · Opportunity AI

Measured against your strategy

  • Together we set a target picture based on your strategic goals.
  • With AI you explore areas you have never been in before, linked to those goals.
  • It can rarely be counted in hours. We measure it qualitatively, against the plan and through interviews, when we hand over.

The endgame · After the handover

What matters most

  • Use of your AI licences stays at the same level as when we handed over. We follow it by group and role, and by person if you wish.
  • Your most active users, the heroes or whatever you choose to call them, keep pushing things forward.
  • The agents are used, corrected and maintained by you, and keep creating value.
  • New agents and automations are built without us.
  • Ambassadors in your organisation have taken over building the technology.

When all of that holds, we know the power to innovate and change stays with you, even as we reduce our focus. That is the endgame.

Example

How we work with Orkla

Orkla is one of our larger clients, and this is exactly how we work with them.

Orkla

We have helped Orkla in many areas, from advice to leadership to agents we built together. Several of them have already been handed over, and there we see that use continues among the participants while our own involvement has gone down.

  • Training
  • Advisory
  • Workshops and masterclasses
  • Building technology and agents

Our philosophy

Use the platform you already have

You have already invested in it, and your data is there. There is no need for an extra data layer on top. Whether you work in Microsoft, Google or something else, you keep your freedom and flexibility, with the least possible complexity and full control of costs.

Your platform · Microsoft, Google or another

Everything stays with you
  • Your data
  • Your licences
  • Your agents
  • Your automations
  • Your knowledge

Not needed

Another layer on top
  • The consultant’s platform
  • An extra data layer

Switching is harder than you think. In a Zapier survey of 542 executives in the US, 74 per cent said the business would be disrupted if they lost their main AI vendor. Almost all of them, 89 per cent, believed they could switch within a month. But of those who had actually tried, 58 per cent said it failed or was much harder than expected (IT Brief, 2026).

The EU has chosen the same direction. From 12 January 2027, cloud providers may not charge a customer for switching provider (European Parliament and Council, 2023, art. 29). The freedom to switch is meant to be the customer’s.

Warning signs

Five signs that your AI consultant is building dependency, not capability

  1. 1. They charge per user for their own platform

    Per-seat licences mean you pay to use what you built together. That is where you should start to worry.

  2. 2. They sell their own version of tools you already have

    If it is the same tool with their label on it, you pay twice. Gartner calls the phenomenon agent washing: existing tools sold under a new label. Of the thousands of vendors selling agentic AI, Gartner estimates that only around 130 offer the real thing (Gartner, 2025).

  3. 3. Your data has to move in with them

    An extra data layer alongside your platform makes it complicated to leave, and you pay for something you already have.

  4. 4. Their framework comes before your problem

    They have the solution ready before they understand your processes. Linthicum describes how ready-made frameworks lead to architecture that is bigger than the problem, for example several cooperating agents where a single model with good instructions would have been enough (Linthicum, 2026). The architecture should grow out of your requirements, not the other way round.

  5. 5. There is no plan for handover

    Nobody in your organisation builds anything themselves, and nobody knows when the assignment is finished. Ask from the start: how will we know we no longer need you?

After the endgame

Want us to stay? Gladly, but in a different way.

Once the capability is in place, you do not need us to build it. Anyone who wants to can keep us on, but at a much lower frequency and intensity.

Then it is about support for leadership: questions about costs, investments and licences, on the board or in an AI advisory group. It is a choice, not a dependency.

Your turn

Now you can choose for yourselves. Where do you want to start?

Everything here is free to try, and you learn something along the way. The questions themselves make you think. If you choose to go ahead, we get a better start together.

If you lead and want to see where you stand

Play the game

Five minutes, five areas and fifteen questions about how you work with AI today. You get a recap of your answers by email, and if you go ahead we start from them in the first conversation.

Play the game →

If you want to experience what employees meet

Try the AI interview

Talk to our AI interviewer, the same one that interviews employees in a mapping. In an assignment, the interviews become the basis for your plan. A trial conversation is not used and is deleted within 90 days. The interviewer currently speaks Swedish.

Try the AI interview →

If you need a figure for leadership

Work out your AI debt

See what it costs not to use AI fully. The calculator runs in your browser, and nothing you type is sent to us.

Work out your AI debt →

If you already know you want to move forward

Book a free meeting

A first conversation costs nothing, and we do not hold back. You leave with concrete ideas, whatever you choose afterwards.

Book a free meeting →

References

Sources