HENUSSE Making your company readable by AI. Outside, and in.

The Operational Method · AI integrations

How do you integrate AI into a real workflow?

HENUSSE starts from the task the team already carries out, or builds a new workflow with them. The tool structures the work, passes it on and makes it visible. The installation is self-hosted in your environment by default. The exit route remains open.

  • First call with no commitment
  • Written scope before any work

The knowledge base

Two modes, one rule

Two modes are available: integrate an existing workflow or build a new one. In both cases, the scope, data and responsibilities are set out in writing.

Two ways of working

Start from what exists, or design a new workflow

01

Integrate an existing workflow

The team keeps the useful action it already takes. HENUSSE connects data entry, structure and the deliverable to remove unnecessary rework.

02

Build a new workflow

When there is no workable process, we scope the stages, data, approvals and rollback before we build.

03

Define responsibility

The tool prepares and passes on. One person remains responsible for validation and the decision that follows.

The underlying issue

An assistant needs the company’s context

An assistant without the company’s context improvises. It lacks your rules, vocabulary, files, offers and past decisions.

The knowledge base documents these rules and this knowledge in a database owned by the company. The context remains readable, retrievable and reusable from one workflow to the next.

HENUSSE connects this knowledge base to work tasks, checks and expected outputs. The knowledge base is designed to scale and remains in your environment: readable, documented and retrievable.

Every new use starts from the same documented knowledge base: your database, your rules, your knowledge and your checks.

Scene 01 · Mobile CRM

From the field to the dashboard, with no re-entry

This scene is the lightweight CRM being built at Protac France, with its marketing director. Léa dictates from her car after a visit. AI structures the sentence, records the fields in the database and updates the director’s dashboard. The scene below shows the complete workflow.

Several business functions, one knowledge base

Context captured once supports several uses

02

Prepare a written quotation

“Prepare the quotation for this client according to our price lists, terms and history.” The salesperson’s notes and internal rules are used to prepare a coherent first structure.

One person approves the scope, amounts and dispatch.

03

Marketing and specifications

“Prepare the label specifications for products X and Y, for this client, according to our rules.” The same knowledge base brings together the product sheet and the applicable rules.

The claims and supporting documents remain subject to validation by the responsible team.

04

Meetings and handover

Meeting notes organise decisions, responsibilities and deadlines. The shared database can then help a new team member find the reference documents.

What is missing becomes visible and can be documented by the right person.

The same knowledge base is used everywhere: sales, marketing, quotations and support. Because your rules and knowledge are captured once, clearly, everything else draws on them. This is where time is saved. That is why the knowledge base is built before the individual uses.

The context is prepared in advance once: onboarding, product rules, terms, history and articles. The assistant retrieves and assembles information from your rules to prepare consistent outputs across business functions.

With a database, rules and named checks, every new use starts from context owned by the company.

Operational sovereignty

Your database, your rules and your history stay with you

The installation is self-hosted in your environment by default. Your database, your rules and your history stay with you. When a workflow calls an external engine (ChatGPT, Claude, Gemini), what is sent to it is specified in the written quotation.

The setup is documented and reversible. Every engagement delivers its as-built file, included and handed over on delivery, as on a building site: where the data lives, what each step does, which rules apply, how to stop everything or hand it to someone else. The handover inventory (data, rules, documentation) is set out in the quotation. And if one day you leave, we can assist the transition, in days at the published rate: we would rather be kept for the work than by the lock.

01

Data

We set out in writing where your data enters, where it stays and who can read it.

02

Validation

We set out in writing who checks the deliverable and who makes the decision.

03

Exit

We set out in writing how to retrieve the data and stop the workflow without losing the knowledge.

The starting point

Choose a workflow worth integrating.

Price

Scoping costs €3,000 excl. VAT. The project starts at €12,000 excl. VAT. The same price for everyone. The full price list is published on the Operational Method page.

The 30-minute call starts from a practical action, a point of friction and the data already in circulation.

WhatsApp