
How Jeausserand Audouard turned its Custom GPTs into CRM copilots with an MCP server connected to Microsoft Dynamics
KEY FIGURES
5 days to deployment
100% of CRM writes validated
Seconds to enrich a lead
CHALLENGE
Jeausserand Audouard manages a dense Microsoft Dynamics CRM: company records, contacts, opportunities, and referred in/out data at the heart of all its business development activity. But for the teams, working with this database day to day remained tedious: manual searches, time-consuming updates, lead enrichment requiring constant back-and-forth between tools, and client relationship mapping that was impossible to maintain by hand.
At the same time, the firm had already adopted Custom GPTs and a ChatGPT Enterprise account for certain tasks, but these tools remained disconnected from the CRM, with no access to real business data.
Before the project began, several key challenges were identified:
- Connecting AI directly to CRM data without bypassing security and governance
- Automating low-value tasks: searching, creating and updating records
- Enriching leads via Kaspr without leaving the work environment
- Mapping the links between clients, companies and opportunities
- Ensuring traceability and validation before any write to Dynamics
SOLUTION
Qolaig designed and deployed a secure MCP (Model Context Protocol) server, connected to Microsoft Dynamics and Kaspr, that integrates directly with Jeausserand Audouard’s Custom GPTs and ChatGPT Enterprise account. The goal: turn the AI the teams were already using into a true CRM copilot, with controlled read and write access to all business data. The same custom MCP approach was deployed at Adelya to connect a sales assistant to product data.
01. An MCP connected to Microsoft Dynamics
The custom-built MCP server exposes CRM features directly within the GPT interface. Teams can query Dynamics in natural language, get summarized records, and create or update CRM objects, all without leaving their work tool. An approach similar to that of Unibail-Rodamco-Westfield, whose teams query their data in natural language.
Examples of real queries:
“What is the latest opportunity recorded for this client?”
“Create a company record for this prospect with the following information.”
“Who referred in this opportunity, and what are their connections with our other clients?”
02. Automatic enrichment via Kaspr
The MCP includes a Kaspr connector that enriches a lead directly from the GPT: the agent searches for available information online, summarizes it, and submits it for validation before writing it to Dynamics. No copy-pasting, no duplicate data entry.
03. Client connection mapping
The agent can navigate the CRM’s relationship graph, identify links between individuals, companies and opportunities, surface referred in/out data, and present these connections as an actionable summary when preparing for a meeting or a deal.
04. Governance and traceability by design
Every write action (creation, update, import) requires validation before execution. Permissions are role-based and every operation is logged: a non-negotiable requirement for a firm handling sensitive data on high-stakes transactions.
RESULTS
- Dynamics accessible in natural language from the firm’s Custom GPTs
- Lead enrichment in seconds, without leaving the GPT environment
- Zero unvalidated writes: full governance and traceability on every CRM action
- Relationship mapping available in real time before every deal
- Deployed in 5 days, reusable across all of the firm’s internal GPTs
RETURN ON INVESTMENT
For a firm whose value rests on the quality of its information and the speed of its relationship execution, every minute saved on CRM management is one more minute spent on the deal. Qolaig’s MCP removes the friction between AI and business data, turning a general-purpose GPT into a true transaction assistant grounded in the reality of the client portfolio. By centralizing access to Dynamics and Kaspr in a single reusable MCP server, the firm now has a durable AI infrastructure that can be extended to new internal GPTs without rebuilding the connectors. As at Dana Avocats, AI fits into the firm’s existing tools without disrupting how people work.
FAQ
How much does an AI workshop with Qolaig cost?
The AI workshop costs €5,000 excl. VAT for one day on site. 50% of this amount is deducted from development fees if the project moves forward.
How long does it take to deploy a solution?
The first use cases are usually live 2 to 6 weeks after the workshop.
Is our data kept confidential?
Yes. It is never used to train third-party models and stays in your environment, with hosting suited to your GDPR requirements.
What did Qolaig deploy for Jeausserand Audouard?
A secure MCP server that connects the firm’s Custom GPTs to Microsoft Dynamics and Kaspr.
How are writes to the CRM controlled?
Every creation or update requires validation before execution, with role-based permissions and full traceability.
How long did deployment take?
5 days, with a server that can be reused across all of the firm’s internal GPTs.

%20(1).webp)


















.webp)



%20(1)%20(1).webp)

.webp)





