01
A map of how the work actually happens
Interviews with the people doing the work, plus a look at the systems and spreadsheets in between. The real process, including the parts nobody wrote down.
← ALL SERVICES
START HERE
1–2 weeks
Agent Opportunity Audit
One to two weeks, a fixed fee, and a written answer: where agents will pay for themselves in your company, where a script is enough, and where people should stay in charge.
- process mapping
- feasibility
- evals
- roi
- gdpr
- eu ai act
WHO THIS IS FOR
Built for you if…
- 01
Operators and founders who know AI should be doing more of their work but can't say which work, or what it would save.
- 02
Teams of 20–200 people where the same approvals, chasing and copy-paste eat a few hours of everyone's week.
- 03
Companies that ran an AI pilot that never left the pilot, and want to know why before trying again.
- 04
Anyone about to sign a large AI build who wants an independent read first.
WHAT YOU GET
Everything included.
02
Ranked opportunity list
Every candidate process scored on hours recovered per week, how risky mistakes would be, data readiness, and build cost. Each one gets a verdict: agent, plain automation, or leave it with people.
03
Feasibility check on your real data
For the top one or two opportunities we run a small, honest test against a sample of your actual documents, tickets or records. You see a measured result, not a slide.
04
Agent boundaries, drafted
For each recommended agent: what it may do, what it must escalate, and what it may never touch. This becomes the spec if you build.
05
Data protection and AI Act notes
Where personal data flows, what has to stay in the EU, and whether any recommended use is likely to fall into a regulated category. We flag it in plain language — your counsel makes the legal call.
06
A written plan with fixed prices
Sequenced recommendations, each with scope, timeline and a fixed price. Use it with us, with another team, or in-house.
07
Readout session
A live walkthrough with whoever decides, including the opportunities we told you not to pursue and why.
WHAT IT LOOKS LIKE
Some of what you get back.
| Supplier invoice matching | 14h/wk | AGENT |
| Customer intake triage | 9h/wk | AGENT |
| Weekly ops report | 4h/wk | SCRIPT |
| Contract approvals | 3h/wk | HUMAN |
Not everything should be an agent. Some steps need a script, and some should stay with a person. The audit says which is which.
- cost / run€0.11
ceiling €0.50
- p50 latency740ms
budget 2s
- daily spend€38
cap €80 → queues
- escalation rate6.2%
tuned weekly
Budgets are code, not a monthly surprise. Crossing the per-run ceiling switches to a cheaper path; crossing the daily cap queues and alerts.
HOW IT WORKS
The process.
01
Kick-off + access
Day 1Agree which teams and processes are in scope, and get read access to the tools and sample data we need. NDA first if you prefer.
02
Interviews + process mapping
Days 2–4Short sessions with the people doing the work. We follow a few real cases end to end and time where the hours go.
03
Feasibility tests
Days 5–8Small tests on your real data for the strongest candidates, so the recommendation rests on a number rather than optimism.
04
Plan + readout
Final daysThe written plan, the ranked list and the boundaries, delivered in a live session. You keep everything whether or not we build it.
TIMELINE & PRICING
What it costs.
DURATION
1–2 weeks
BUDGET RANGE
€2.5k – €5k
Typically €3.5k
PAYMENT
50% to start · 50% at readout.
WHERE IN THE RANGE YOU LAND
- ·€2.5k: one team or one process family, one feasibility test.
- ·€3.5k: where most land — two or three teams, two feasibility tests, full written plan.
- ·€5k: wider scope, messy or scattered data, or a regulated workflow that needs more care.
- ·Credited in full against any Elicon build that starts within 60 days.
- ·Fixed fee agreed before we start. No surprise extension.
COMPANY-WIDE
Several departments, multiple sites, or a regulated environment needing deeper data-protection work is a larger discovery, scoped the same way: fixed fee, written plan, and credited against what follows.
STACK
What we build it with.
- Interviews
- Process mapping
- Sample evals
- Cost modelling
- GDPR review
NOT INCLUDED
Where the line is.
- ×Building anything beyond the small feasibility tests — that is the next step, quoted in the plan.
- ×Legal advice. We flag data-protection and AI Act questions; a lawyer answers them.
- ×Change management or staff training programmes.
- ×Vendor selection for software you would buy rather than build (we will say if buying is the better answer).
FAQ
Questions we get.
- Isn't this just a sales call you charge for?
- No. You leave with a written plan, measured feasibility results and fixed prices that any competent team could execute. Plenty of audits end with us recommending a small script, or nothing at all.
- What if the answer is that AI isn't worth it for us?
- Then that is the answer, in writing, with the reasoning. It is cheaper to learn that for €3k than after a €30k build.
- How much of our team's time does it take?
- Usually two or three hours per person interviewed, plus someone who can grant access and answer follow-up questions.
- Do you need access to our live systems?
- Read access to a sample is enough. We work on exports or a sandbox where possible, and nothing leaves the EU.
- Is the fee really credited?
- Yes, in full, against any build that starts within 60 days. If you build with someone else, the plan is still yours to use.
BUDGET · €2.5k – €5k
Not sure where agents fit yet?
Tell us which teams feel most stretched. We'll come back with what an audit would cover and a fixed fee.
Book the audit →