Streamline D365 Finance company creation with a Copilot Agent

Introduction

In the context of ERP projects, the ability to quickly have coherent and realistic demonstration data is a key factor for accelerating workshops, validating processes, and training teams. The manual creation of a company and its basic data (clients, suppliers, articles, parameters, sequences, taxes, accounting…) often takes several hours—or even several days — and involves rare functional experts.

Creating a Copilot Studio agent capable of automatically generating a complete company from scratch in D365 Finance & SCM brings a new dimension of automation: a guided, conversational, reproducible and audited experience. The agent becomes a real ERP provisioning assistant that transforms a tedious and technical operation into a simple natural interaction.

This article explains how to configure a new legal entity and add new demo data to Dynamics 365 F&O using the Dynamics 365 ERP MCP server and the Business Process Catalog from an agent in Copilot Studio. Previously this process could take weeks; now it can be done in one or two hours.

Installation

I first downloaded the Business Process Catalog from here.

I then created a new Copilot Studio agent called “D365 Finance Create Legal Entity” and selected the model “Claude Opus 4.6”.

I then uploaded the Business Process Catalog files as knowledge sources and added the Dynamics 365 ERP MCP server and the Microsoft Learn Docs MCP server to the agent.

In the instructions, I used the standard instructions for the Dynamics 365 ERP MCP server, simply adding the following section at the end:

Full instructions:

# Role

Act as an autonomous data entry agent responsible for interacting with the Dynamics 365 Finance and Operations app.

There are 3 types of tools for interacting with Dynamics 365 Finance and Operations applications: form tools, API tools, and data tools.

– Form tools enable interaction with F&O forms in the same way a user would through the UI.

– API tools allow calling custom X++ logic.

– Data tools allow interacting with F&O using OData.

# Tool selection guidance

– For create/read/update/delete operations – you MUST prefer using data tools before using form tools.

– When explicitly instructed, or if proved impossible to complete the task using data tools, use the form or API tools.

# Objective

Your objective is to respond to tasks provided by the user. First execute each step of the provided task workflow using your MCP tools. Check if you have achieved your objective after each tool call. If you have not achieved your objective then continue to execute the next step in the task workflow.

# Form tool Usage Instructions

– Typical flow for record creation operations is to find a menu item, open a form, click the new button, find and set values for relevant controls, save the form.

– You can use grid filtering to find relevant records for update, delete or inquiry scenarios.

– DO NOT EVER EVER ask for menu item types. The find_menu tool groups menu items by their type.

# Data Tool Parameter Filling Instructions

– Get the entity types and their schema metadata using data_find_entity_type tool before using CRUD data tools, unless instructed otherwise in the task instructions.

– You MUST use plural entity name in the OData path for data operations. For V2+ entities, use plural before V. E.g. SalesOrderHeadersV2.

– DO NOT use deep insert (nested entity creation in one call), it is not supported.

– For filtering by enums (or setting enum values), use the following format: $filter=Style has Namespace.Pattern’Yellow’.

# Form Tool Parameter Filling Instructions

– Omit optional parameters if no value is provided as input.

– Use menu item names (and not labels) when filling menu name parameters in tool calls.

– Use grid column names (and not labels) when filling grid column name parameters in tool calls.

– Use control names (and not labels) when filling control name parameters in tool calls.

– Use tab names (and not labels) when filling tab name parameters in tool calls.

– (lessThanDate(x:int)) is a valid value for a grid date column filter.

# Extraction Instructions

– A tool call response can include up to 25 rows of data as form state. Generate a warning if the form state contains 25 rows of data.

# Reasoning Instructions

– Think out loud and reason step by step.

– Before each tool call, plan the action.

– After each tool call, reflect on the result and determine the next step.

– When answering questions about data DO NOT rely on your general knowledge. Use tools to find accurate and precise data

– When instructed to create new data, and the creation fails, DO NOT retrieve existing data instead.

– DO NOT stop reasoning until all tasks are complete or an error prevents continuation.

– DO NOT stop reasoning to ask a user questions or ask for user input.

– Only ask questions if the task is not clear

# Knowledge Source

– Primary: Microsoft Learn Docs MCP Server

– Secondary: Dynamics 365 ERP CMCP Server Schemas

– Additional: Uploaded knowledge sources:

— Business Process Catalog Tree (Why) – Full DEC 2025.1.xlsx

— Success by Design Delivery Plan DEC 2025.xlsx

— Deliverables (What) Tree DEC 2025.1.xlsx

Creating a legal entity

Configuration plan

I used the following prompt:

Create a configuration plan for creating a new legal entity in Dynamics 365 Finance and Supply Chain Management.

Use the added knowledge base files to follow the required configuration steps.

The plan should contain all steps required to configure the General Ledger, Accounts Receivable, and Accounts Payable modules. The new company should be named ‘MCP1’ and be based in Lyon, France.

Use everything that has already been implemented that could help configure the new company.

We should later be able to execute the steps one by one.

Please make the plan downloadable so that I can refer to a specific step and execute it in a new chat session with this agent.

It even included a recommended execution sequence at the end.

I copied it into a text file and saved it on my computer.

I could then split it much more easily for Copilot.

Company Creation

All I had to do was:

Start a new session
Upload the plan and instruct it to execute the different steps in the plan:

Execute SECTION 1: FOUNDATION SETUP

Based on the attached deployment plan Execute Section 1: FOUNDATION SETUP (Prerequisites)

Execute SECTION 2: GENERAL LEDGER CONFIGURATION

Execute SECTION 3: TAX CONFIGURATION (FRANCE SPECIFIC)

Execute SECTION 4: ACCOUNTS RECEIVABLE CONFIGURATION

Execute SECTION 5: ACCOUNTS PAYABLE CONFIGURATION

Execute SECTION 6: BANKING CONFIGURATION

Execute SECTION 7: CURRENCIES AND EXCHANGE RATES

Execute SECTION 8: INTERCOMPANY CONFIGURATION (If Applicable)

Execute SECTION 9: VERIFICATION AND TESTING

I could also ask the agent what had already been completed.

In one case, I had to start a new chat session, upload the plan again, and ask it to resume execution from the step where it had stopped (internet connection loss).

Demo data

For the demo data, I used the following prompts.

Vendors

In legal entity MCP1 create 10 new vendors; they should all be production companies of car spare parts.

Products

Create 20 new products and release them to legal entity MCP1.

All items should be car spare parts.

Add 2 approved vendors to each product using the existing vendors.

Purchase orders and lines

MCP1 needs inventory to sell.

Create purchase orders for all released product items.

Choose vendors that are approved vendors for the items.

Please add 2 PO lines to all open purchase orders.

Select only items where the vendor on the purchase order is an approved vendor for the item.

Set cost and sales prices that align with market norms.

Customers

Create 15 customers, all being car repair stores.

Give them unique names while keeping the car repair store theme.

Fill in as many fields as possible so they can purchase items from us.

Sales orders and lines

Create 2 sales orders for different customers.

Add 2–3 sales order lines to each of them.

I could continue like this with all the data I would like to generate!

Business advantages

1. Massive time saving in projects:

Drastic reduction of the time needed to create a demo company: from several days to a few minutes.
Reduction of manual work for the basic configuration.
Freeing up consultants’ time for higher-value tasks.

2. Standardization and data quality:

Systematic generation of consistent, complete and business model-compliant data.
Reduction of human errors in initial configurations.
Alignment with corporate templates (CoA, dimensions, taxes, workflows…).

3. Acceleration of workshops and POCs:

Possibility to re-generate a demo environment at will for each workshop.
Almost immediate creation of specific scenarios: Retail, Manufacturing, Distribution, Projects…
Improvement of the customer experience through adapted and personalized demonstrations.

4. Empowerment of business teams:

Non-technical users can request the creation of a company by simple conversation.
No need to wait for an IT intervention or an ERP consultant anymore.
Acceleration of business prototyping.

5. Project cost reduction:

Decrease in days/man dedicated to initial setups.
Optimization of the schedule: less waiting for the preparation of environments.
Free mutliplication of test scenarios.

6. Improvement of the time-to-value:

Teams can start working earlier in the project cycle.
Decisions are informed because they are based on realistic data.
Customer feedback comes sooner, better responsiveness.

7. Facilitation of automated tests:

More reliable regenerable data for test scenarios.
Possibility to recharge a fresh company before each campaign.
Alignment with Data Setup oriented CI/CD pipelines.

8. Multi-scenarios and multi-local support:

Generation adapted to countries (taxation, formats, purchase/sales book).
Modular configuration according to the business, volume or parameters.
Reuse of prompts for several clients or BUs.