Skip to content

Services — AI & automation

AI and automation that removes real work

We apply AI and automation to the manual handovers, inboxes and spreadsheets that slow your team down, and we evaluate it before it ships.

01 — Overview

Software that sits in the workflow, not a demo on the side.

Most of the value in AI and automation is unglamorous: an order that flows in without re-typing, a message that lands with the right person, an issue that is routed on its own. We look for the places where that kind of work piles up and build around them.

We pair models with plain rules where rules are enough, and keep people on the exceptions and the decisions. The aim is less admin, not a system nobody trusts.

AI och automation i praktiken

When teams ask for AI och automation, the real question is usually which manual step to remove first. We start there: map the flow, pick one step, automate it, and measure whether the work actually went away.

02 — What we do

What you get.

  • Workflow automation

    Orders, requests and issues flow in digitally and are routed by rules, so nobody re-types or chases.

  • AI-assisted triage and drafting

    Models that read, tag and prioritise incoming work, and prepare drafts that a person approves.

  • Integration with your systems

    Automation connected to the systems you already run, so it fits the day instead of adding another tool.

  • Evaluation and monitoring

    AI features are evaluated before launch and monitored after, with a clear hand-off to a person.

03 — How we work

How we approach it.

  1. 01

    Follow the work

    We walk the real flow with the people who run it and find where time and mistakes pile up.

  2. 02

    Start small

    One step, automated first, so you can see the effect before you widen the scope.

  3. 03

    Keep humans in charge

    People stay on the exceptions and the decisions, with the full history in front of them.

  4. 04

    Measure, then expand

    We track a small number of agreed goals in the open and extend only what is working.

  • Daily communication with the people doing the work
  • Ship, measure, refine — before you scale the bet
  • Room to pivot when the data says the first idea was wrong

04 — Related work

Where we have done something similar.

Issues are categorised and routed to the right contractor automatically, and the property team steps in only on exceptions.

A bright open workspace with people working at laptops and meeting in a glass-walled room

Real estate

A property owner with residential buildings and a wish to do things better than the industry norm.

Automated issue handling for the property owner who wanted to innovate.

Tenants report an issue from their phone with a photo. The system sorts it, sends it to the right contractor and keeps everyone updated until it is fixed. The property team now steps in on exceptions instead of every email.

  • Strategy
  • Product Design
  • Automation
  • Development

Read the full story

Requests, handovers and returns moved from email and spreadsheets into one app.

A woman working on a laptop at a bright office desk, with colleagues and storage bins on shelves behind her

Workplace operations

A growing company keeping track of laptops, screens, furniture and everything in between.

Track all your office inventory in one place. No more emails and spreadsheets.

We built one place to register, assign and find every item the company owns, from laptops to office chairs. Requests, handovers and returns happen in the app, so the spreadsheet and the email threads could finally retire.

  • Product Design
  • Development
  • Automation

Read the full story

05 — FAQ

Questions we hear often.

Where does AI actually pay off?
Where there is a steady stream of repetitive, text-heavy or rule-based work: sorting messages, routing requests, preparing drafts, moving data between systems. We would rather automate one such step well than promise a transformation.
Will AI replace our staff's decisions?
No. In our work AI supports judgement and does not replace it. Routine steps are automated, and people approve anything that matters and handle the exceptions.
How do you make sure it works before launch?
We evaluate AI features against real examples from your own workflow before they ship, and monitor them afterwards, so problems show up early.
Do we need to replace our existing systems?
Usually not. We aim to connect to the systems you already run and automate the handovers between them.

Which manual step should go first?

Describe how the work runs today. We will tell you where automation or AI would take the pressure off, and where it would not be worth it.

  • Reply within two working days
  • Straight answer from a senior lead
  • No obligation