PAI-DOC-06
2026-07 · Offers
An ERP copilot prepares your management actions. A human validates them, always.
An ERP copilot is an AI connected to your ERP that prepares your management actions: quotes, proposals, orders, entries, built on your real customer, stock and order data.
Every action is presented ready to check, only goes out after human validation, and leaves a trace: the system never runs fully autonomous.
It is our Régie offer, the third level of the PropulAI dossier: it includes team training and the private AI over your documents.
01
An ERP copilot does not simply answer: it executes.
The ERP copilot is the third level of the PropulAI dossier, our Régie offer, the complete package: everything the private AI brings, connected to your ERP. The AI no longer only reads your documents, it reaches your operational data: customers, stock, orders, in real time.
It prepares your management actions end to end: a complete quote, a proposal, a supplier order, ready to check and send.
The limit is set from the start: every action passes through human validation, and the system never runs fully autonomous.
02
The flow fits in one diagram: prepared by the machine, validated by a human.
Customers
accounts, history
Stock
availability, prices
Orders
current, past
The copilot
- ERP reading
- Preparation
- Justification
The action goes out
Validatedquote sent, order placed, entry posted
It comes back
Returnedadjusted or refused, with the reason recorded
Customers
accounts, history
Stock
availability, prices
Orders
current, past
The copilot
- ERP reading
- Preparation
- Justification
The action goes out
Validatedquote sent, order placed, entry posted
It comes back
Returnedadjusted or refused, with the reason recorded
Fig. 01 · The copilot prepares, a human validates, every action leaves its trace.
The copilot reads your ERP data, prepares the action and presents it with its justification: the documents and figures that explain it. A human validates, adjusts or refuses, and the action only goes out after.
The rule is the same as for our whole practice: the machine reads, reconciles and alerts; decisions stay human.
03
The first actions prepared are the ones that cost you hours.
The quote is the textbook case: the copilot crosses the client request with your prices, your stock and the account history, and prepares a complete, costed proposal, to proofread rather than to build.
Then come the everyday actions: pre-sales answers on up-to-date data, supplier orders prepared when a stock threshold is reached, entries reconciled with their documents.
The exact scope is defined after the diagnostic, action by action: each automation is priced per unit before it is connected.
04
Governance is reinforced where processing is sensitive.
Every prepared action is recorded: which data was read, what was proposed, who validated, when. That audit trail makes the system presentable to an accountant or an inspection.
Sensitive processing (prices, payroll, personal data) receives reinforced governance: written rules, dedicated validations, systematic logging.
That is what the EU AI Act expects of professional AI use: governed, demonstrable practices, produced by the very way the system works.
05
The copilot is built on a proven foundation, never first.
Régie includes the previous levels: team training and the private AI over your documents. We do not connect a copilot to an ERP before the foundation has proven itself at your premises.
The installation follows our five-stage engagement protocol: Diagnostic, Scoping & compliance, Pilot, Deployment, Managed operations. You decide at each step, on figures measured at your premises.
We are paid for the project, never for reselling licences: the choice of components follows your situation, not a catalogue.
06
We connect systems to ERPs coming from our own workshops.
Our agency has been building ERPs and business platforms for SMEs since 2024: more than fifty sites and platforms delivered, ten internal vertical SaaS products in production. The copilot connects to systems we know from the inside.
DocAgora, our e-health platform in Portugal, operates with a supervised agent in a stricter setting than day-to-day management: health data, human oversight, logging.
Detailed case studies are available on request, during the diagnostic.
07
Continue
08
Commitments
Four commitments, on every engagement
These commitments are set out in our presentation deck. They hold for every engagement, from the diagnostic to managed operations.
- Contact
- Dedicated, end to end
- Progress point
- Weekly
- Environments
- Permanent access
- Reply
- Within 24 working hours
09
Decide
Two minutes to know where you stand.
A straight answer: what applies to you, and whether a project makes sense. If the answer is no, you leave with the answer.
Regulation (EU) 2024/1689 · PAI-DOC-06