Skip to content

Two systems that finally talk to each other.

Orders coming out of the webshop and typed into the ERP by hand. Stock levels that are never right. Accounting that only knows what was sold at the end of the month. As long as your systems stand apart, someone is closing that gap every day.

We build the connection between them. Also when one side has no decent API, and also when the system dates from 2011.

What it costs you every week.

Companies run on dozens of applications of which fewer than a third talk to each other. People fill that gap, every single day.

What we tie together.

Usually one side is something modern and the other has been standing there for twenty years. Both are fine.

  • ERP and accounting

    Exact, AFAS, Twinfield, SAP, Odoo. Orders in, invoices out, without anyone sitting in between.

  • Webshops and marketplaces

    Shopify, WooCommerce, Magento, bol and Amazon. Stock, prices and orders in both directions.

  • Warehouse and logistics

    WMS, carriers, packing slips and track and trace. Customs documents included if you cross the border.

  • CRM and marketing

    Customer data that is right in one place, instead of three lists each claiming something else.

  • Machines and meters

    Production lines, sensors and inverters. What happens on the floor ends up in the system.

  • Systems without an API

    Then we read what is available: a database, an export folder, a file format from 2009. Less elegant, works fine.

How we build a connection.

More than eight in ten integration projects fail to deliver what was promised. Almost always because the old side was underestimated or because it was never tested with real data.

Book a call
  1. We follow one order

    From purchase to invoice, by hand, through every system. That shows exactly where something gets retyped and which fields appear or vanish along the way.

  2. One direction, with real data

    We start with the flow that saves the most work and test it with your own data, not a tidy sample file. That is where the exceptions surface.

  3. Alongside the manual work

    The connection runs for a while as people still do it themselves. If the numbers match, the manual work goes. If not, nothing is broken.

  4. We keep watching

    Every connection we build reports itself when something goes wrong. Failed messages are kept and sent again, and we hear about it before you notice.

A connection will break at some point.

That is not pessimism. A supplier changes their API, a key expires, someone fills in a field that was never there. The question is not whether it happens, but whether you notice before your customer does.

So we build around it as standard: a log you can read yourself, messages that try again, and an alert to us when something stays stuck. That sits in the maintenance, we do not invoice it separately.

Questions about connecting systems.

Not answered here? Just ask. You get someone who builds these connections themselves.

Ask your question
Our system has no API. Is it still possible?

Almost always. Then we look for another way in: straight onto the database, a folder where exports land, or a file the system writes out every night. Less pretty than a real API, but from where you sit the only difference is that it works.

What does a connection cost?

A straightforward connection between two systems that both have a decent API is a matter of weeks. Add an old system, or a lot of exceptions, and it climbs. After the first call you hear a figure per flow, so you can start with the connection that returns the most.

How long does it take?

The first working flow is usually there within two to four weeks. A complete project with several directions and systems runs closer to two or three months. You do not have to wait for everything before the first piece of work is off your plate.

What if our supplier changes their API?

Then we hear about it, because the connection reports itself the moment something stops adding up. Small changes we absorb within the maintenance. If a supplier throws out their whole API, that is a separate job and we say what it costs beforehand.

Does it run at your end or ours?

Whichever you prefer. Usually the connection runs on our servers in the Netherlands, because then we can watch it. If you want it in your own environment, we put it there and write down how you manage it.

What happens to data that does not get through?

It stays put rather than disappearing. A failed message is retried automatically, and if it stays stuck you see it in the log and we get an alert. Nothing quietly falls on the floor.

Can you connect to something we built ourselves?

Yes, and that is often easier than a closed package. We read your code, or we build an API on top of it so you can hang other things off it later as well.

Who owns the connection?

You do. The code, the keys and the servers are in your name. If you continue with another party later, you hand it over without anything staying behind with us.

Got a question?