AI consulting

AI consulting with fixed scope and named deliverables.

Advice billed by the hour has no reason to end. Our engagements have a deadline, a scope agreed before we start, and a document you keep at the end — so you decide how you want to proceed.

Book a call (30 min) →
How it is set up

An engagement, not an open-ended retainer.

AI consulting in practice often means a series of workshops, a deck of recommendations, and a contract that renews until someone decides to stop. It is hard to say when the work is finished, because the end was never defined.

Every engagement here has three things written down before it begins: how long it runs, what is in scope, and what you hold at the end. A monthly retainer exists, but it is optional and it comes after delivery — not instead of it.

Deliverables

What you actually receive.

Each service stands on its own. Take one, two, or all three as a single engagement. Where you start depends on how much you already know about where the problem is.

Diagnose · 2–3 weeks

Measurement before building. The output is a document with numbers about your business — usable whether or not you work with us afterwards, and whether or not someone else implements it.

Design · 1 week

The decision on what to build, what to buy, and in what order — made on feasibility and impact, not on the order the ideas arrived in.

Deliver

The plan carried through to a working system, including the people who will use it. The full scope is on the services page.

3D Bundle

For organisations that would rather not make three separate decisions and seek three separate budget approvals.

What a finding looks like

A finding is a number, not an impression.

The difference between an analysis that gets used and one that sits in a drawer is whether its claims can be checked. So every finding carries a person, a step in their day, and a time estimate. Two examples from our own case studies:

An hour a day, one person

A service company in Belgrade: the same data retyped into three systems after every call. The finding was not “there is room for automation” — it was how many hours a week, and whose.

Two to three days per event

A policy organisation in London: every event started from a blank page. Multiplied by the number of events a year, the decision to invest stops being a matter of judgement.

Independence of the recommendation

We do not live off licences we would sell you.

When an adviser also sells the licence, “buy this” is never a fully neutral recommendation. We do not resell licences or generic platforms, so build-versus-buy has no predetermined answer.

One possible outcome is that we build on a subscription you already hold — Claude, ChatGPT, Gemini or another provider. We advise which model suits your use case, build the automation on it, and train your team on what was built. No new vendor, no new infrastructure.

Where the diagnostic points to a platform, the work can be delivered on Integral — hosted by us, or entirely on your own servers. Where the system runs after handover is your decision, not our assumption.

When to call

Three situations where this makes sense.

You cannot see where the problem is

You know time is going somewhere, but you have no number. Start with the diagnostic; nothing further is required afterwards.

You have more ideas than capacity

Proposals are everywhere. Design ranks them by feasibility and impact, and gives you the order to do them in.

You need proof before a wider commitment

Pilot runs the whole loop on one process in six weeks, against a success criterion written down before the work starts.

Common questions

Frequently asked.

What does your AI consulting actually cover?
The base form is the diagnostic — measurement before building. The output is not a deck of recommendations but a document with numbers about your business: where the time goes, whose it is, what it is worth, and in what order to address it. Exactly what is included is set out on the services page.
Is this a monthly retainer?
No. Diagnose, Design, Pilot and the Delivery Sprint all have fixed scope and a known end date. A monthly retainer exists as an option after delivery, covering a monthly strategy call, minor adjustments and priority support. You take it only if you want it.
Can we take only the diagnostic?
Yes. The diagnostic is a fixed-scope engagement of two to three weeks and requires no further commitment afterwards. The document you receive is yours, and you can act on it alone, with us, or with someone else.
We already have someone else’s analysis. Can you work from it?
We can. Design accepts opportunities identified through our diagnostic or brought in from your own. If diagnosis and design are already done, the Delivery Sprint builds one or two specific automations in two to four weeks and hands them over.
Who on our side needs to be involved?
The diagnostic means interviews across the whole team, not only management, because inefficiencies are tied to specific people and the steps they carry out. It also needs time from someone who knows your systems, for the systems and data-flow analysis.
What if you conclude that we do not need AI?
That is a legitimate outcome and it is what the document will say. If the right answer is something else entirely — a different process, a simpler integration, or no change at all — we will tell you that. You are paying for the analysis, not for a predetermined conclusion.

Start here

Fixed scope, known end.

On a thirty-minute call we go through where you think the time is going, and the systems that hold that work together. From that we propose a scope — a diagnostic, or design straight away if you already have the analysis. The document that comes out of it stays with you whatever you decide next.

Book a call (30 min) →

or write to ai@meteorit.rs