Andreas AhoniemiHead of Digital, Gullström & Co
SvenskaBook a call

AI automation

AI automation means a workflow is executed by software instead of a person, with language models handling the steps that require interpretation: reading an email, understanding an invoice, routing a ticket, drafting a reply. I build these flows in n8n and Python, connect them to the systems you already run, and leave logs behind so every execution can be audited afterwards.

How the flow runs

  1. The trigger

    An email arrives, a form is submitted, the clock hits seven.

  2. The interpretation

    The language model reads and works out what it is about.

  3. The checks

    Code sums, looks up and validates. Never the model.

  4. The approval

    A person confirms anything hard to undo.

  5. The execution

    The system updates, the reply goes out, everything is logged.

Where it usually pays off

Ticket routing, inbound quote requests, invoice processing, product data for e-commerce, report assembly and internal knowledge search. What they share is variable input with a rule-driven decision, which is precisely the gap language models fill.

Humans in the loop

Automations that decide without visibility get switched off within six months. So every flow ships with an approval step where one is warranted, alerts when something deviates, and a full log of what the model saw and what it answered.

What it costs not to

Before anything gets built we measure the current state: how often the process runs, how long it takes and what it costs in salary. If the numbers do not work, we do not build it. Reaching that conclusion in a meeting is cheaper than reaching it after a project.

Start with an hour

Bring a process that grates. We walk it together and you get a straight assessment of whether automating it is worth doing.

Book a call