Capture what your best people know. Automate what nobody should have to do.
In a discovery of ten working days we capture how your people do their work. What is knowledge, we build into skills and plugins the rest of the team works with. What is routine, we automate. You get a roadmap with a cost estimate and only then decide whether we build.
You will hear where we would start, and where we would not.
Seven steps. You decide only after the roadmap.
From a thirty-minute call to automations that take work off your hands. The discovery is the only paid step before you decide whether we build.
01 · Intro call: Honest advice: discovery or not
02 · Proposal: In writing before we start
03 · Discovery: Five days with your people
04 · Roadmap: Order, impact, cost estimate
Decision point: Here is where you decide.
05 · Setup: Your rules, your tenant
06 · Build: Person-days per month, with your people
07 · Measure & maintain: Measured per milestone, then maintain
Honest advice: discovery or not
30 min
We listen to where things get stuck for you and say whether a discovery makes sense. If it does, we fix the scope in the same call: which teams, which processes. If not, you will hear that too.
In writing before we start
Scope · teams · timeline
Which teams we talk to, which processes we look at and when. So you know in advance who we take off the floor and for how long.
Five days with your people
€6,000 · 10 working days
Five days of interviews with the people who know the work best, on site or online: how they do it, which exceptions they know, what is written down nowhere. Then five days of analysis.
Order, impact, cost estimate
Included in the discovery
What we build and in which order. Per item we estimate how many days the build takes and what impact we expect. Plus three paces: steady, standard, fast. So you know roughly what the build costs before you decide.
Here is where you decide.
Paid up to here: €6,000
After the roadmap you choose: build or not, and at which pace. Up to here you pay only for the discovery.
Your rules, your tenant
Your tenant · Copilot Studio or Claude
We start from your policy: who may do what, where it runs, what stays inside the tenant. Result: a working environment, agreed with IT.
Person-days per month, with your people
8 person-days per month (standard) · estimate per item in your roadmap
The roadmap in milestones of skills, plugins, connectors and automations, built on what your people told us in the interviews. Every month a piece goes into use; we work their feedback into the month after, until it works.
Measured per milestone, then maintain
KPI report per milestone · feedback every month
Per milestone we measure what we fixed in advance. Then it keeps running: maintenance, monthly feedback interviews, an agreed response time. When someone joins or leaves, we update the skills.
“So you end up with another roadmap.”
What you may be thinking now
True, if it stops at a document. With us, that is where the work starts.
- A plan we carry out
The roadmap is not advice you have to deliver yourself. We build what is in it, in that order.
- Knowledge stays in-house
How your best people work then lives in skills and plugins their colleagues work with. Even after they leave.
- Agreed in advance, measured afterwards
Per milestone a KPI we fix before the build.
- We carry the adoption
Every month a feedback round with the people who work with it. That it gets used is our job, not just that it exists.
One figure before we build, one figure after.
In almost every discovery the same question comes up: how do we know afterwards whether it delivered anything? Like this: per KPI three measurements, on the same line in your roadmap.
- Now
4hours
Half a day. Measured in the interviews, with the people who do it.
- Target
1hour
In the roadmap, per milestone. The figure we aim for.
- Measured
After milestone 2, in the KPI report. The same measurement, again.
An ambitious target, not a contractual promise.
What you still want to know after this page
About money, time, IT and what happens if it disappoints. The rest you will hear in the first call.
That is in your roadmap, as an estimate per item: how many build days and what impact we expect, with a cost estimate and three pace scenarios. At the standard pace, ArrowEye works around eight person-days per month; the monthly price is agreed after the roadmap, based on that pace. Before the roadmap we deliberately name no amount: that would be guessing. The discovery itself costs €6,000, and until the decision point that is all you pay.
Siemen — co-founderNo, and we say so in the first call. If you already know exactly which process you want to automate and it is clearly scoped, ArrowEye tells you whether a discovery of ten working days makes sense or is overkill. Usually it does: the interviews capture what is written down nowhere, and precisely that determines whether a skill works afterwards. If the answer is no, you hear that too.
Maxime — co-founderFive days of interviews, spread across the people who know the work best, on site or online. Per person that usually means one or two conversations of an hour to an hour and a half, not five full days of one colleague. The five days of analysis afterwards are ours. Which teams and processes we look at is settled in the first call, so you know in advance who we are taking off the floor.
Siemen — co-founderThen you know after ten working days instead of after a year of building. The roadmap states honestly what ArrowEye would not build and why: too little impact, too many exceptions, sources that need to be put in order first. The roadmap is yours, even if you take it to another party or pick it up internally. The decision point comes only after that; until then you commit to nothing.
Maxime — co-founderYes. ArrowEye then builds in Copilot Studio and Power Automate, within your own Microsoft 365 tenant, with your security groups, your SharePoint permissions and your DLP policies as the boundary. The setup happens before the first build, together with IT. If Claude is allowed, we choose per item what fits best; if it has to be Microsoft, that is the frame, and we say up front what can and cannot be done within it.
Siemen — co-founderBecause we define it up front. For each milestone we agree a KPI and measure the situation before we build: how many hours something takes, how long a request is in transit. Before go-live, every skill runs through an evaluation set of real questions from the team. After the milestone we measure again and you receive the report. The target is ambitious, but it remains a target, not a contractual promise.
Maxime — co-founderIn the maintain phase, ArrowEye does: monthly feedback interviews with the users, an agreed response time and skills that are updated when someone joins or leaves. If you want to do it yourself, that is possible too: at delivery your IT receives a handover document, a maintenance guide and the evaluation set, so they can test every change themselves.
Siemen — co-founderYes. The build runs month by month, in milestones, and you pay per month. If a milestone does not deliver what the roadmap predicted, we first adjust based on the users' feedback. If you still want to stop, you can do so whenever you want: what has already been built runs in your own environment and is handed over with a handover document and a maintenance guide.
Maxime — co-founder