Do we have to replace our software to use AI?

The goal is to take manual work away, not to add a tool nobody wants to use. A migration on top turns a two week project into a track with no end date, and then the work it was all about disappears from view.

So we build on what is there. That is also why most teams need a short explanation and not a training programme: they keep working in the screen they were already looking at.

The mailbox, the document folder, the accounting package, the ERP, the planning board, the CRM, and the spreadsheets that in practice still sit next to all of it. From Microsoft 365 to Notion: we build applications around the tools your team uses.

Through an API where there is one, and otherwise through the way in that does exist: an export, a database or a file that is prepared every night. We check per system which connections are available and build on those.

Code that reads and writes on those systems, with an identity of its own and access no wider than the job. Read from this one folder, write to this one table, send from this one mailbox, nothing else.

And a log. Per run it records what came in, which steps the system took and what went out. Without that log an automation is a black box, and nobody trusts a black box for longer than two weeks.

The keys sit in the environment and never in the code, and they can be replaced without anything breaking. If the route needs an extra action later, that is a change someone approves before the access gets wider.

We do not sell an off the shelf connector to a package. Which connections a system offers differs per package and per version, so we look that up per assignment before we give a price.

If there is no usable way in, we say so, and then that is the outcome of the call rather than a surprise halfway through the build.