Skip to content

Work and systems

For controlling departmentskontroling

In a management report, each number has its source.

Data for the report comes from several systems. The system reports according to established definitions and records where each number comes from. A comment with a number without a source is not published.

Tested on sample data, without customer data.

numbers without a source in the test reports
0
comments with the number of unsourced comments stopped before publication
4/4
data sources in one report
3

Syntalith test results on fictitious data from 24 editions of the report. Last checked: October 22, 2026.

Closing the month means copying the numbers from exports.

Data comes from the accounting system, CRM and spreadsheets. The team manually reconciles the definitions and checks whether the comment matches the latest version of the data.

  1. 01

    One definition

    The same measure calculated in the same way every month.

  2. 02

    Source of the number

    Each number opens its source.

  3. 03

    Visible differences

    Differences between systems do not disappear with reconciliation.

From data to report in four steps.

  1. 01

    Downloads data

    From the accounting system, CRM and controlling sheet.

  2. 02

    Agrees

    Compares values and shows differences with cause.

  3. 03

    He submits a report

    The source is written next to each number.

  4. 04

    The controller approves

    Checks for exceptions and comments before publishing.

What the system does and what stays with the controller.

Makes a system

  • It collects data from three sources
  • Shows the differences between systems
  • Stops a comment with a number without a source

The controller decides

  • Measurement definitions
  • Explanation of unusual deviations
  • Publication of the report

Statistics: how much does it cost to close the month manually in 2026.

According to the Syntalith test (2026), there was not a single number in the test reports without a source.

All 4 comments with an unsourced number were stopped from publication.

Preparing the commentary and checking the numbers cost about €0.0010 in the test.

According to the Central Statistical Office, the average salary in the enterprise sector in September 2026 was PLN 9,266.48 gross. With employer contributions and 168 hours a month, an hour of work costs the company PLN 66.45.

Scenario: one month close

By hand
80 hours of work, PLN 5,316
AI comment
approx. €0.0010 for preparation and inspection
He stays with people
definitions, deviations and publication

Respect. Assumption: 80 hours of work at the end of the month is an example number. We measure the time after implementation at the client's.

What's left with the controller.

Publication
The report is published only after approval by the controller.
Definitions
The team determines how to count each measure.
Deviations
An unusual difference awaits human explanation.

This is how the controller works.

Screens from the test using fictitious data.

  1. Otwórz raport miesiąca
    01Otwórz raport miesiącaKontroler zaczyna od gotowego raportu.
  2. Sprawdź, skąd jest liczba
    02Sprawdź, skąd jest liczbaKażda liczba ma źródło.
  3. Wyjaśnij różnice między systemami
    03Wyjaśnij różnice między systemamiRóżnice nie znikają z raportu.

Is your team also gluing together a report from several exports?

Describe your process to us. We will show the prototype on your process and your data. The analysis and prototype cost €0, and you know the fixed implementation price before the contract.

What our OpenAI and Anthropic partnerships mean for you.

OpenAI makes ChatGPT and Anthropic makes Claude. We belong to the partnership programs of both companies, so we work with their tools every day and learn from the creators themselves.

OpenAI Select PartnerClaude Partner Network
01We will select a tool for the task
We implement both ChatGPT and Claude. We check both on your documents, and you choose the one that makes fewer corrections and saves more time.
02Knowledge straight from the creators
We learn from official OpenAI and Anthropic courses and materials. Your team learns about the features and settings that work today.
03Licenses without our margin
When we roll out ChatGPT, Claude or Copilot, you buy the licences directly from OpenAI, Anthropic or Microsoft and they invoice you. We add nothing to the licence.
04Certified engineers
On the team, we have an AWS certification in machine learning and an Anthropic certification in systems architecture on Claude.

Participation in programs does not mean that OpenAI or Anthropic endorses our services.