ENGAGEMENT / 02 · ADVISORY
Know what to build before you spend a cent building it.
Four to six weeks of pure thinking work. We map what you already run, score how ready you actually are, rank every AI opportunity by return, and hand your board a roadmap it can sign off.
- 4–6
- WEEKS
- 2
- EXEC SESSIONS
- 1
- RANKED ROADMAP
- 18K
- FROM, IN EUROS
01 / What this actually is
What this actually is.
Advisory is head work, not code. You are buying the undivided attention of a senior Lumacode team that spends four to six weeks inside your organisation, talks to the people who run the processes, and comes back with a decision document instead of a demo.
By the end you know three things with numbers attached: how ready you really are, which use cases pay back first, and what the first twelve months should cost. Everything we produce is yours, in writing, in a format you can hand to a CFO.
WHAT IT IS NOT
It is not a pilot, not a prototype, not a proof of concept. Nothing gets built in this phase — that is deliberate, and it is what makes the next phase cheap.
02 / Who this is for
WHO THIS IS NOT FOR
If you already know exactly what to build and who will use it, skip this and start with an implementation partnership.
03 / How it runs
How it runs.
01 · WEEK 1
Kickoff and system inventory
Interviews with the business units and with IT. What systems exist, what data lives where, which processes people actually follow versus the ones in the handbook.
02 · WEEK 1–2
Maturity assessment
We score you against the things a working AI system needs: data quality and access, existing tooling, in-house skills, governance and compliance posture. Scored, not described.
03 · WEEK 2–4
Opportunity map
Every candidate use case in the business, plotted on effort against value. Quick wins, big bets, distractions and traps get named as such.
04 · WEEK 3–5
Technology evaluation
For the shortlisted cases: build, buy or wait. Which platforms and models fit your stack, your data residency rules and your budget — and what each one really costs at your volume.
05 · WEEK 5–6
Executive workshop
We present the roadmap to your leadership, defend the numbers, and leave with decisions rather than a follow-up meeting.
03B / The opportunity map
Every idea, on one grid.
Each use case gets plotted on effort against value. The argument stops being about opinions and starts being about position.
QUICK WINS
High value, low effort. These start in month one.
BIG BETS
High value, high effort. Sequenced, staffed, gated.
HOUSEKEEPING
Low value, low effort. Do them if they are free.
TRAPS
Low value, high effort. Named so nobody restarts them.
← EFFORT → · ↑ VALUE ↑
04 / What you hold at the end
What you hold at the end.
05 / Who you work with
Who you work with.
01
Senior strategist
Runs the interviews, owns the business case and presents to your leadership.
02
Technical architect
Judges what your stack and your data can actually carry, and prices it honestly.
03
Governance lead
Checks each opportunity against EU AI Act obligations, data protection and internal review.
06 / Price
from 18,000 €
FIXED FEE · 4–6 WEEKS · FIXED SCOPE
INCLUDED
- All interviews, workshops and analysis
- Every document listed above, yours to keep and reuse
- Two executive sessions, on site or online
- Thirty days of follow-up questions after handover
NOT INCLUDED
- Any build work — that is a separate engagement
- Licences or cloud spend you decide to take on
- Ongoing operation of anything you later deploy
07 / Questions
Questions we get.
How long does it really take?
Four to six weeks for most organisations. Very large groups with many business units run to eight; we say so before we start, not halfway through.
Do we have to build with Lumacode afterwards?
No. The roadmap is yours. You can run it with your own team, with us, or with someone else entirely. We would obviously like to be there, but nothing in the deliverables is locked to us.
Who from Lumacode is in the room?
A senior strategist and a technical architect throughout, so the business side and the engineering side are answered in the same meeting rather than in sequence.
How much of our time does this cost?
Roughly six to ten interviews of forty-five minutes, one half-day workshop and two executive sessions. We work around your calendar, not the other way round.
What if the honest answer is 'do not do AI yet'?
Then we write that down and tell you what to fix first — usually data access or process ownership. We would rather lose the build than sell you a system that cannot work.
08 / Where this sits
Where this sits.
The cheapest part of an AI programme is the six weeks you spend deciding what not to build.