<aside> 💫 Goal: Provide instructions to the Karomia AI on which data points to regenerate and how to approach this.

</aside>

Introduction

Based on the details and documents you provided during the report creation process, the Karomia AI will initially generate as many answers as possible for the data points within your report scope.

Once the process is complete, you’ll be notified and can review the output. If needed, you can provide instructions on how and where improvements are required, and the Karomia AI will regenerate the affected data points accordingly.

<aside> ℹ️

The initial run might not meet expectations, as responses may miss relevant company context or key documents may be missing. To improve automated results in the second AI run, we include this review phase.

</aside>

Steps

Step 1: Navigate Your Report

After opening a report, you’ll see its structure on the left. Your report is organized into two levels:

1. Topics: ESRS 2 General Disclosures; E-topics; S-topics; G-topic.
2. Disclosure Requirements: The thematic breakdown of each topic as specified in the regulation, e.g., G1-2 – Management of relationships with suppliers.

Only topics and disclosure requirements with data points included in your report scope will be displayed. Others are omitted. You can collapse the navigation menu to give yourself more space to review the data point answers for a specific disclosure requirement.

An example of how the structure of your report will be shown on your Karomia platform

Each listed disclosure requirement contains one or more data points, displayed on a single, scrollable page. You can access the page by clicking on the specific disclosure requirement in the report structure menu.

You’ll encounter different types of data points:

1. Narrative: Requires a textual answer, with formatting options available.
   • Example
2. Semi-narrative: These can include:
   • Booleans (True/False)
     • Example
   • Single-select (choose one answer from a list)
   • Multi-select (choose one or more answers from a list)
     • Example
3. Numerical: Requires an integer or decimal value, often with a specified unit.
   • Example
4. Table: Requires answers organized into a predefined table structure.
   • Example