Sensei

In every company, someone just knows.

Which invoice to approve. Which one to hold. And why. Sensei captures that judgment while experts work, has them confirm it, and teaches it to the next person.

Open app
One-minute demo · sound on
Why it matters ↓
THE PROBLEM

When she retires, her know-how walks out the door.

Most of it was never written down. It lives in the heads of people who have done the work for years, and leaves when they do.

The manual

What to do.

“Software invoices are approved by accounts payable.”

The expert

Why.

€180 software invoice. Approve? “It depends.” On the project, the month, the budget.

HOW IT WORKS

From one expert to every new hire.

観1 · watch & ask why

Capture

Sensei sits quietly beside the expert while she works in her usual tools. Each click becomes one screenshot that AI reads. When she pauses, a voice agent asks why.

“Is that €100 limit per invoice, or per project, per month?”

型2 · rules, confirmed

Work Map

Her answers become rules, each linked to the moment on screen and her own words. Sensei reads the rules back in under a minute, and she confirms them.

IF Claude spend > €100 per project per month → THEN the project lead approves.

教3 · guide new hires

Teach

A tutor watches the new hire on real cases. Before a wrong decision is saved, it stops her, explains why with the expert’s words, and lets her fix it herself.

€90 invoice + €40 already spent = €130 → send to the project lead.

観Why?Sensei asks in the language the expert thinks in.
WHY TEAMS USE IT

Know-how that compounds.

No forms. No extra meetings.

Experts keep working. Sensei asks a few questions at natural pauses and offers a draft to confirm. Nobody has to write documentation.

Captures the why, not just the what

Manuals say what to do. Sensei captures the conditions, limits and exceptions experts check before they decide.

Every rule has a source

Each rule links to the screen moment and the expert’s exact words, and the expert confirms it. Nothing is invented.

Mistakes stopped before they happen

The tutor reviews a decision before it is saved, not after the payment run. New hires learn from a near miss, not a real one.

Any language

Warum, pourquoi, ¿por qué?: experts answer in the language they think in. The rules work for everyone.

The expert controls the record

Screenshots only after clicks, never while idle. Off the record pauses screen and microphone. A help corner asks Sensei only when you want it.

THE KNOWLEDGE MODEL

Every decision, broken into seven dimensions.

Context, baseline, condition, rationale, authority, release criteria, and the evidence behind them. Sensei asks about each one and marks what is confirmed and what is still unknown.

Explore the knowledge dimensions →
  1. ABusiness context
  2. BExpected baseline
  3. CDecision condition
  4. DBusiness rationale
  5. EAction and authority
  6. FRelease criteria
  7. GEvidence and validity
THE BIGGER PICTURE

Every expert. Every decision. One knowledge base that keeps growing.

Each answer adds a piece. Confirmed rules, expert quotes and screen moments build up into knowledge that never walks out the door.

Today

New colleagues

learn the why from day one, on real cases.

Tomorrow

AI agents

act on the same confirmed rules, with the expert’s reasoning attached.

HOW IT’S BUILT

Works beside the tools teams already use.

Browser

Expert at work

Works as usual in the ERP, with the tab shared.

Claude Haiku · ElevenLabs

Capture

One screenshot per click, read by AI. A voice agent asks why at pauses.

Claude Opus

Work Map

Answers become rules, with quotes and screen moments. The expert confirms.

ElevenLabs agent

Teach

A tutor watches the new hire, stops a wrong save and explains why.

Sensei, teach the next generation.

See Sabine’s judgment captured and passed on to Lena in one minute.

Open app