Running ten apps and don't want to change your ERP?
We can still help you with AI.

Odoo is our first choice because it gives a company one system for everything. But not everyone can switch right away. So your apps can stay: we connect them to Odoo as a shared data foundation and add AI on top. Native, third-party, or local on your own server.

We recommend Odoo first.
But it isn't the only path to AI.

We propose what makes sense for you. Neither of the other two paths closes the door on moving to Odoo later.

02

Odoo as the data foundation

Your ERP and apps stay. We connect their data to Odoo, where it is linked and unified. Reports and AI then run on top of it.

  • Connected via API or regular exports
  • Customer, job and document in one place
  • Move areas into Odoo step by step

Best when replacing your ERP isn't an option right now, but you want your data together.

03

Local AI over your current data

The model runs on your server or in a dedicated cloud, and we connect it directly to your systems. Odoo isn't required.

  • Data stays under your control
  • Read-only access to selected data only
  • We solve a specific task, not a generic chatbot

Best for sensitive data and a clearly defined task. Requires suitable hardware.

Your apps stay.
Your data meets in one place.

AI connected to each app separately only ever sees part of the company. When data first meets in Odoo, AI works with the whole picture: the same customer, the same job, the same numbers.

Your apps

Current ERP
CRM
E-shop
Manufacturing and inventory
Excel and documents

Odoo

Customers and jobs
Products and documents
Permissions and history

AI

Native in Odoo
Third party
Local model

More than data storage

Odoo isn't just a warehouse where data lands. People can work in it right away, and every connected area is a step towards a unified system.

If you later decide to retire an app, part of the journey is already behind you. If not, the connection simply keeps working.

01

A ready-made business data model

Customers, products, jobs and documents already have their place and links in Odoo. We aren't building a data warehouse from scratch.

02

Permissions and traceability

Everyone sees only what they should. AI can be limited to selected data, and you can trace what an output is based on.

03

People work with the data directly

Reports, tasks and approvals run in the same system, not in yet another app on top of spreadsheets.

04

A natural path to a unified system

When the company wants to replace an app, that area moves into Odoo, where its data already is.

Native, third-party or local?
The task and data sensitivity decide.

Native AI in Odoo

Where the model runs
With the provider you configure in Odoo
Best for
Text, emails, summaries and questions right in Odoo
Keep in mind
Available features depend on version and edition

Third-party AI

Where the model runs
With the model provider, e.g. ChatGPT, Claude or Gemini
Best for
More demanding tasks and custom agents over your data
Keep in mind
Data goes to the provider, so we settle which data and on what terms

Local model

Where the model runs
On your server or in a dedicated cloud
Best for
Sensitive data and repeated, well-defined tasks
Keep in mind
Hardware and operation. A smaller model suits narrower tasks
More about AI inside Odoo →

We connect it to the data you already have

This isn't a chatbot you upload a few documents to. We connect the model to your current systems, databases and documents, with read-only access to only what you choose.

For questions about numbers, the model doesn't guess. It turns the question into an exact calculation, which an application runs over your data. So the result can be verified, and the model doesn't need to see the whole database.

🖥️ On your server
🔎 Read-only access
🧮 The application does the maths
🗂️ Databases and documents
🔗 Later on top of Odoo too
Illustrative example

“Which customers' revenue fell the most this year compared with last year?”

Model understandsRevenue decline, this year vs last, by customer
CalculationRun by the application directly over your data
Sent to the modelOnly the text of the question
AnswerAn overview linked to specific documents

We adapt the approach to the task and data sensitivity

With one process, not a big project

  1. 01

    We map your apps and data

    Where data is created, which system is the source of truth for what, and what is sensitive.

  2. 02

    We verify one process in a pilot

    We pick a task with a clear benefit and show on your data whether it works.

  3. 03

    We connect more sources

    We extend what proved itself and set up permissions and output review.

  4. 04

    We decide what's next with Odoo

    Based on the results, we suggest which areas are worth moving to Odoo and which to leave alone.

Straight answers

Why not move to Odoo completely right away?

If you can, we'll recommend it. One system is the cleanest long-term solution. But sometimes it doesn't make sense right now: you've recently rolled out an ERP, a big project is running, or you want to see the benefit on a smaller part first. Then we start by connecting, and you decide about switching later.

Will we have to move to Odoo afterwards?

No. The connection works on its own. Switching is an option, not a condition.

Is Odoo really a data warehouse?

Not in the sense of a large analytics platform over huge data volumes. For everyday business data such as customers, jobs, documents or inventory, though, it's usually enough as a shared foundation. If your volumes or analytics need more, we'll say so up front.

What if our app has no API?

We can also connect via the database, regular exports or files. In the analysis we check what is reliable and how current the data will be that way.

Does local AI make sense if our data is in the cloud?

Usually less so. If data already sits with cloud services, a local model only addresses part of the concern. It makes the most sense for sensitive data in your own systems, or when you don't want to send it to another provider.

Is local AI as good as ChatGPT or Claude?

For general questions and complex reasoning, usually not. For a narrowly defined task, it can be enough. That's why we first verify it on your data and your hardware.

How much does it cost?

It depends on the number of connected systems, data quality and running the model. You get the scope and price after the analysis, before we start anything.

Want to know which path is right for you?

We'll go through your apps and data and tell you straight whether Odoo makes sense as the main system, as a data foundation, or whether local AI fits. No strings attached.