AI strategy & consultancy
Know what to automate. And what to leave alone.
A one-hour session and a short audit, ending in a written map: what to automate first, what to leave, and what a spreadsheet already fixes. The map is yours, even if you build with someone else.
Verdict
Automate first
High volume, and the rule is writable: days overdue sets the reminder, and one system holds the record. This is the shape of work an agent is actually for.
Runs on·n8n · your accounting API · a human sign-off over an amount you set
Pick a job
Why start here
The costly mistake isn't a bad build. It's building the wrong thing well.
Most AI projects do not fail on the technology. They fail because the wrong thing got built, confidently, before anyone checked whether it was worth building. The assessment is where you check.
Which of these ideas is real, and which is just hype?
Where do we start, and what is any of it actually worth?
Is our data and our process ready, or would we automate a mess?
How do we avoid spending on the wrong build?
The workshop
It starts with an hour, not a form
Before any audit, one session. We map the business from the inside, because a system that does not fit how you work is a tool, not a solution.
Where the week goes
The repeated work, named task by task, so the audit measures real time instead of a guess. This is where most of the automatable money is quietly hiding.
Where it stops sounding like you
The points where a customer meets a generic reply, a slow follow-up, a form. Small frictions, but they are what your brand is actually felt as.
What you are building toward
Where the business wants to be in a year. The map recommends what serves that, not simply what is technically possible to automate.
The deliverable
You leave with a written map
Not a slide deck of possibilities. The finished map: every job on your list, its verdict, and the next step beside it. A short, plain document you can act on the same day, or hand to another firm.
Strategy map
prepared for your business
n8n + your accounting API
a linked cell you already have
write the rule down first
your call; numbers on tap
Yours to keep, whether you build with us, with someone else, or not at all. It starts with the hour.
Who it's for
Built for the person who has to decide.
You are weighing whether AI is worth it at all, and you want a straight answer, not a pitch.
Your team is drowning in repeat work, and you need to know which parts are safe to hand to an agent.
You need a roadmap you can defend upstairs, not a pilot that quietly stalls.
What it's worth
Grow the output. Not the payroll.
Think of each agent as one hire's worth of the repeat work. A working ecosystem is usually three to five of them.
In salary terms
CHF168’000
a year in salary you do not add as you grow
Each agent carries about one hire's worth of the repeat work.
Built once, owned by you: no salary, no recruitment, no ramp.
The agents carry that load without the salaries, running in your own accounts. Your people spend their hours on judgment and brand, not retyping.
Book the hour.
We run the session, you keep the map, and nothing gets built until it is the right thing.
FAQ
Before the audit
What this stage answers, and what it does not.
A readiness audit scores you against a checklist. This one starts from your week and your brand, then names specific work to automate, to leave alone, or to fix, with the reasoning written next to each call. You can act on it the same day, and it never assumes the answer is to build something.
Then that is the finding, and you still keep the map. Sometimes the honest answer is a setting in software you already pay for, or a process that has to be written down before any agent can touch it. Saying so costs us the build; not saying so would cost you more.
No. The map is yours to keep and to hand to another firm if you would rather. It is written to be read by anyone, not coded to lock you to us. If you do build with us, the audit is the process the build needs anyway, so nothing is repeated.
Four things, in plain terms: how often the work repeats, whether the rule behind it can be written down, whether one system holds the record, and what a mistake costs. Work that is high volume, rule-writable, single-source, and low blast radius is where an agent pays for itself. The audit scores each candidate against those and shows its working.
About an hour of the people who actually do the work, and read access to see where the time goes. The written map follows within days. Doing the audit does not commit you to a build, and it is not a retainer: it has a scope, a price, and an end.
No. The map names candidate tools per job, from the same menu a build would use: models like Claude, GPT or Gemini, orchestration like n8n or LangGraph, and whatever you already run. It picks against your job, not our habits, and it says why.