Half-day
Foundation
The whole team together. Demystify, secure, and walk away with the AI Standard Operating Procedure in hand.
Bring my team up to speed →AI Training for Businesses · Switzerland
Train your team on AI without exposing your data
"The Augmented Employee" brings the rigor that transformed manufacturing to your office: a standard. Your team practices on your own use cases and walks away with the habit of protecting your data on every request.
10+ businesses trainedFrom 2 hours to a full dayOn-site or by video call
Why this training exists
Lean eliminated waste on the shop floor. The office kept its own.
An operator does not guess the next move: they follow a written, continuously improved standard. In the office, nobody has a standard. Everyone improvises their request to AI, gets an approximate result, and starts over.
Worse: one employee in three is already using public AI without leadership knowing. Contracts, payroll, client files, manufacturing specs. Shadow AI is not a threat on the horizon — it is already happening.
We are not here to add one more tool. We are here to set the standard that is missing, and give your team the habits that protect you.
Signature training formats
The Augmented Employee
Choose based on what you want to fix first, with 1 to 12 people per session. Whichever format you pick, your team walks away with the same standard in hand. Prix sur demande.
Full dayRecommended
Rollout
Foundation, plus workshops split by department. Each team works on its own files. The format for a first serious rollout.
Organize my rollout →2 to 3 hours
AI Sprint
One department, its real files, a measured gain by the end. Short, focused, repeatable. Tailored by role: an AI Kaizen Sprint on the shop floor, an AI Close Sprint at an accounting firm, an AI Front-Desk Sprint for government offices.
Launch a sprint →2 hours
Leadership Session
For your leadership team. What you allow, what you forbid, who decides. Walk away with the three decisions to make before your team makes them for you.
Bring my leadership team together →The results
What your company gains, starting the following Monday
- Your data no longer leaves by accident: everyone knows what never leaves the company, and how to anonymize a document before submitting it.
- Hours of drafting time back every week: a complete request on the first try, using the Standard’s five blocks, instead of four rounds of rewording.
- Not one more invented answer reaches a client: the team spots a fabrication in three seconds and demands the reasoning, never the conclusion alone.
- Office waste disappears: double entry, re-keying, copy-paste, waiting on a file. Shop-floor Lean, applied to the back office.
- The decision stays with you: at the first sign of doubt, the AI stops and calls in a manager. It proposes, your team decides — responsibility never gets delegated to a machine.
- AI stops being two curious employees’ side project: a role-based assistant gets written once, usable by the whole department, and handed off to the next hire.
These gains are not taken on faith. The workshops run on your own files, never on demo examples: you see the result before the session ends.
Who leads it
Led by a production engineer, not an IT consultant
Victor Semblanet, founder of Palambur, based in La Chaux-de-Fonds. A Mines Nancy engineer and Lean expert (Kaizen, 5S, VSM). An SAP migration in energy, then a production manager role in luxury jewelry: machining, polishing, setting.
The examples come from real files, not textbook cases. And AI is treated as a production tool: it demands a standard, just like a machine tool does.
« I don’t sell technology. I sell time back, risks removed, and lasting competitiveness. »
The content, module by module
Six modules, then the workshop
01
Vision and industrial reality
The Golden Triangle, Lean 4.0 and digital waste, the Jidoka of data. And vocabulary that finally makes sense: LLM, tokens, hallucination, memory.
02
Cybersecurity and governance
Shadow AI, the silent eavesdropping of meeting note-takers, and the fix: agents whose data stays physically in Switzerland. Here the attack travels through language — your firewalls never see it.
03
Technical foundations
What is under the hood, explained in shop-floor language. The Transformer as a text production line, attention station by station, and the multi-model approach: the right model for the right task, never a single-vendor dependency.
04
The Palambur Method
Six rules, from putting the instruction first to the anti-hallucination safety net, through to iteration — the Kaizen of the prompt. A prompt is a process: it improves like a machining routine.
05
Use cases by role
Human resources, shop floor and production, administration and finance, engineering. Your roles, your examples.
06
Roadmap and ROI
What you launch on Monday, what you measure at 30 days, and which task you hand off tomorrow at 8:00 a.m.
The core of the method
The Palambur Template: your AI Standard Operating Procedure, 5 blocks in this order
The Template is that same standard, applied to a request made to AI. And one simple principle: a missing block means an approximate result.
Learning by doing
Twelve exercises, on documents that look like yours
As in industry, you only master a tool by using it. Every exercise is done Template in hand. Four of them, to set the tone:
- ManufacturingVoice dictation in a noisy environment turned into a structured report
- ManufacturingThe exact paragraph found within 500 pages of ISO standards
- Human resourcesAn employment contract audited against the applicable labor agreement
- GovernmentIllegible meeting notes converted into decisions and owners
The tools
The tools on the market, and which ones deserve your data
The real question is not "which tool is best." It is which data you hand to which tool. The reference table your team walks away with:
| What you are handling | What you should use |
|---|---|
| Public text: a job posting, a post, a web page | Any mainstream assistant |
| Internal, non-sensitive: a procedure, meeting notes | A business account, with the "do not train on my data" option turned on |
| Personal data, client files, payroll | Redaction required before sending, or a Swiss-hosted solution |
| Manufacturing secrets, specs, patents, contracts | A sovereign agent: data stays in Switzerland, the model is never trained on yours, access is traceable. Or nothing at all. |
Who it is for
Leadership, managers, and frontline teams
Manufacturing businesses (watchmaking, microtechnology, precision machining), local government agencies, accounting and law firms, HR teams, and other regulated professions. No technical background required.
AI governance
What the training leaves behind
A successful session sets up a clear decision loop: who approves a new tool, who rules on a borderline case. And it starts four assets that belong to you:
- Prompt library: the workshops produce your first documented prompts, ready to share across the department.
- Skills: the role-based assistants built in session stay with you and get reused without reinventing the prompt.
- AI Charter: the anti-Shadow-AI ground rules covered in training become the basis of your written charter.
- Design system: the first templates come out of the workshops, then the full set gets formalized during ongoing support.
To scale these assets company-wide: Audit & Advisory →
Open source
The method, open source
What you learn in training is also published in the open. The palambur-method skill applies the same logic step by step: you can read it, install it, and adapt it, before or after your session.
F.A.Q.
Foire aux questions
How long does the training take?
Foundation runs a half-day, Rollout a full day with workshops split by department. AI Sprint and the Leadership Session run two to three hours.
Do I need to already know AI to take part?
No, no technical background required. The training starts from where participants actually are, including teams who are new to AI.
How many people per session?
1 to 12 people, depending on the format: the whole team at once, one department at a time, or your leadership team.
Where does the training take place?
On-site or by video call, anywhere in French-speaking Switzerland. We come to you — your team does not travel.
Is the training just for manufacturing?
No. The method comes from industrial Lean and applies as-is to an accounting firm, a local government office, or an HR department. Examples are adapted to your line of work before the session.
What happens after the session?
A 30-day check-in is included, to answer your questions and fine-tune the habits your team picked up on real cases.
Is our data exposed during the training?
No. Exercises run on anonymized documents, and redaction is the very first habit taught.
Next step
Book a date
Tell us about your business, and we will get back to you within 48 hours with a program tailored to your team and the right format.