Skip to main content
Orki MCP • CSRD • ESRS • VSME • EcoVadis

Your ESG reporting, tracked from your AI assistant

  • Data collection under control: data entered, data missing and contributors running late, framework by framework.
  • Consistent data: Orki's consistency rules applied to your data, with the discrepancies explained.
  • Reporting prepared: ESRS, VSME or EcoVadis datapoints read as they stand in Orki, so you can draft faster.

See the platform in action

An Orki expert will get in touch to arrange your demo.

OK

Request sent!

We have received your request. Our team will be in touch very shortly.

How did you hear about us?

📢 Advertising
🔍 Google
💼 LinkedIn
💬 Word of mouth
🤖 ChatGPT or another AI
❓ Other
Error

Something went wrong

Please try again later or email us directly at contact@orki.green.

Your AI assistantOrki connector enabled

Where do we stand on data collection for the VSME report? List what is missing, by contributor.

Orki · ESG data collection status

VSME 2025 data collection: 46 of 59 datapoints filled in (78%).

ContributorFilled inMissing
Human resources97
Purchasing64
General services112
Finance200

On the HR side, what is mostly missing is headcount by contract type and workplace accidents. Would you like me to prepare a reminder for each contributor?

Fictitious conversation, sample data. The collection status is the ESG module's; the assistant only presents it.

In practice

ESG reporting, in conversation

Requests that an assistant connected to Orki's ESG module can handle (fictitious data).

You

Where do we stand on data collection for the VSME report? List what is missing, by contributor.

Your assistant, via Orki

46 of 59 datapoints filled in (78%). HR: 7 missing, including headcount by contract and workplace accidents. Purchasing: 4. General services: 2. Finance has filled in everything.

You

Are there any inconsistencies between our headcount and our training hours?

Your assistant, via Orki

2 discrepancies. 46 training hours per employee against 9 last year: the total seems to have been entered for all sites combined. And an average headcount (212) higher than the year-end headcount (198) with no departures declared.

You

Prepare a first draft of the paragraph on energy consumption, based on our data.

Your assistant, via Orki

“In 2025, total energy consumption amounts to 4.8 GWh, 38% of it from renewable sources, down 6% on 2024.” Each figure links back to the Orki datapoint it comes from.

You

Which topics come out as material in our double materiality assessment?

Your assistant, via Orki

4 material topics: climate change (E1), on both impact and financial materiality; resource use and circular economy (E5) and own workforce (S1), on impact materiality; business conduct (G1), on financial materiality.

The assistant's tools

What your assistant does in the ESG module

The assistant sees the ESG module of your workspaces within the limits of your ESG role.

Read

Your ESG data, as it stands in Orki.

  • Datapoints by framework (ESRS, VSME, EcoVadis)
  • Raw data and its source
  • The double materiality assessment

Check

What is missing and what does not add up.

  • Progress of data collection
  • Consistency rules and their discrepancies

Prepare

The drafting work, based on your data.

  • First drafts of texts based on the saved values
  • Year-on-year comparisons

Understanding MCP

A standard connector between your AI and your Orki data

You don't need to be a developer to use it. Here is what it is, and what happens when your assistant works in Orki.

Youask your question in plain language
a conversation
Your AI assistantClaude, ChatGPT, Le Chat, Copilot...
the MCP protocol
The Orki platformyour data, your calculations, your permissions

What is MCP?

MCP (Model Context Protocol) is an open standard published by Anthropic in late 2024 and since adopted by most AI assistants. It defines a single way to connect an assistant to a piece of software.

It is often compared to a USB-C port: the same connection works whatever the assistant, and whatever the tool on the other side.

What difference does it make?

Without MCP, your assistant only knows what you paste into the conversation. With the Orki MCP, it reads your data on the platform, runs calculations, checks a file before import and, if you allow it, saves data.

You stay in the tool you already use, with the model you have chosen.

Who does the calculations?

Orki. The assistant never makes up an emission factor or a result: it calls the platform's tools, which apply the same rules and calculation methods as the application.

A result obtained through the assistant is therefore identical to the one you would read in Orki.

Security and control

Your assistant never has more permissions than you

The Orki MCP acts on behalf of the signed-in user, with exactly the permissions that user has in the application.

1

Sign-in with your Orki account

The assistant connects through OAuth, the standard for connections between applications. You sign in on Orki, never in the assistant, and no password is ever shared with it.

2

Your permissions, nothing more

Entities, roles, contributor restrictions, locked workspaces: anything you cannot do in the application, the assistant cannot do either. An auditor stays read-only.

3

The workspaces you choose

When you connect, you choose which workspaces the assistant can see. A workspace you have not ticked remains invisible to it.

4

Write access, on authorisation

Read access is granted by default; write access is turned off. You enable it workspace by workspace, and you can withdraw it whenever you like.

5

Revocable at any time

Connected applications are managed from Orki's settings: one click is enough to cut off an assistant's access.

6

No extra tokens to pay for

The model runs on your subscription to the assistant. Orki provides the data and the calculations, not the model.

Getting connected

Compatible with the main AI assistants

Claude, ChatGPT, Claude Code, Cursor, Mistral's Le Chat, Copilot in VS Code: any MCP-compatible assistant can connect to Orki. An Orki expert opens your access and sets up the connection with you, on your own workspaces.

ClaudeChatGPTClaude CodeCursorLe ChatCopilot

The Orki MCP on other modules

Still have questions?

Everything you need to know about the Orki MCP and how to use it day to day.

Book a demo
Those of Orki's ESG module: the CSRD's ESRS, the voluntary VSME standard and the EcoVadis questionnaire, as well as frameworks specific to your workspace.
It can prepare first drafts from the data saved in Orki, which you then review. The report and the published values remain your responsibility.
Yes. The assistant reads the data at the time of the question, directly on the platform. It sees what you would see in the application.
No. In Claude or ChatGPT, you connect it from the settings by pasting an address and signing in to your Orki account. After that, everything happens in conversation, in plain language.
Any assistant that supports remote MCP servers: Claude, ChatGPT, Claude Code, Cursor, Mistral's Le Chat and Copilot in VS Code, among others. Availability of custom connectors depends on the plan you have with each provider.
Orki only sends your data to the assistant you have connected, and only when that assistant calls a tool. How the assistant then processes it is governed by its provider's terms and your plan: the enterprise plans of the main assistants generally exclude training on your conversations.
No, the two coexist. The built-in agent remains available in the application. MCP lets you work from the assistant you already use, with the model of your choice, and cross-reference your Orki data with your other tools.
Orki platform users, with their own account. Book a demo or talk to your Orki contact: we will show you how it works on your use cases and open your access.

Your ESG reporting, in conversation

In the demo, we connect an assistant to an ESG workspace: data collection status, consistency checks and first drafts of texts.

Book a demo