---
title: "Most of it is already built. We build the rest with you, around your systems and the way you work."
nav: How we work
summary: "How an engagement runs from the first interview at a subsidiary to a first version in your IT environment, which capabilities we bring with us, what is shaped with you, and why AI proposes while people approve."
kind: page
status: written
cta: Talk to an expert
ctaTo: booking
heroCta: none
jump:
  - { label: Three steps, href: "#steps" }
  - { label: Our capabilities, href: "#capabilities" }
  - { label: Your capabilities, href: "#yours" }
  - { label: Your environment, href: "#facts" }
  - { label: Customer cases, href: "#cases" }
related:
  - /people
  - /intelligence
doc: docs/44-formagorna-i-stallet-for-omradena.md
---

We bring knowledge, experience and proven building blocks of our own, and shape the rest around your needs, in your IT environment, together with the people who do the work today. We start by looking at what their day-to-day work involves.

::::band{tone="light" stack}
## Three steps {#steps}

::figure{name="three-steps"}
First a few subsidiaries, then all of them. First one requirement, such as Scope 3, then the next on the same foundation.
::::

## Our capabilities {#capabilities}

What we bring to every engagement. Select a capability to read more.

::figure{name="capabilities"}

## Your capabilities {#yours}

Your part grows over time: your rules, your definitions and every decision with its source.

::figure{name="layer-mark" part="7"}

:::columns
### What we shape with you
What each row represents for you, which codes and definitions should apply across the group, and where the results should go. Which system the data is taken from and which one it ends up in. Which subsidiaries are outside the main systems and how they report. Who approves what, and who can see the details. How far you want to go now, and what can wait.

### What stays as it is
Your systems, subsidiaries, countries and ways of working. The data is collected where it already is, so nobody needs new logins or forms. The specialists make their judgements where they already work, and the group sees its totals where it does today, unless you choose otherwise.
:::

:::dark
## AI reads. People approve. Over time both get better. {#ai}

This is how AI works in everything we build, shown here in [The Cards](https://thecards.eu), our sustainability reporting tool. The assistant proposes a value for a data point from the organisation's own documents and earlier reports, and the proposal points to the file and page it came from, so that the reviewer opens the source instead of relying on a summary. A proposal is saved only when a person chooses to use it, and only a person can mark a value as done. The history of each data point shows whether a version was written by AI, written by AI and then edited or written by a person.

::figure{name="cards-report"}
A report in The Cards: Copilot asks for values through MCP, the assistant proposes them with file and page, and a person accepts. Copilot's access token only lets it propose. The values shown are made up.

Every decision is carried into the next version of the rules: the thresholds, what counts as flagged and the words your organisation uses for its own purchases and facilities. The proposals improve each year, as there is more of your own material to draw on, and people spend their time on judgement calls rather than retyping.
:::

## From the source to the report, in your own environment. {#facts}

::figure{name="data-flow"}
Where the data originates, what runs in your IT environment, such as Microsoft Azure, and where the data is used.

The solution can be maintained by us under contract, by your own team or by another supplier. Everything built for you is in your environment, and for the components we bring with us you get full rights to use and modify them, and to have others maintain them.

## Customer cases {#cases}

:::case{href="/cases/peab" image="/images/cases/peab-e45-goteborg.webp" alt="The E45 in Gothenburg one evening, with the light trails of the traffic" label="Customer case · Peab"}
### Expert judgement, scaled to thousands of accounts
We trained a language model on judgements from Peab's experts and the LCA consultants at Miljögiraff, which then handled the remaining thousands of accounts. Peab now has an AI tool of its own that can be audited and becomes more reliable over time.
:::

## Ready to start?

You can start in several ways: a solution built with you in your own environment, [The Cards](https://thecards.eu) off the shelf, a code or platform review, or support and advice for what you already run.

::cta
