DealCockpit™ and AI: our vision, our commitments, our conviction
We are living in unique times. AI is popping up everywhere – sometimes useful, often heralded with great fanfare. In the world of corporate finance, where every decision involves substantial sums of money, teams of people, and strict legal responsibilities, we have chosen to set out in black and white what we do, what we will not do, and why.
This document is not a sales pitch. It is a clear statement of our position.
Our conviction
Since 2012, DealCockpit™ has been supporting professionals in M&A, private equity, restructuring and corporate finance with the secure management of their data rooms. Our focus has always been the same: to make the complicated simple.
AI is the natural next step in this approach. Not a radical change, not a publicity stunt. A suite of tools that allows our clients to ask questions in natural language, to find information in a single sentence rather than ten clicks, to prepare for an investment committee meeting more quickly, and to ensure no critical due diligence questions are overlooked.
Notre choix : MCP, le standard ouvert
We have opened up DealCockpit™ to AI assistants via MCP (Model Context Protocol), the standard championed by Anthropic and OpenAI to enable language models to interact with business applications using natural language.
This is both an architectural choice and, above all, a philosophical one.
We don’t dictate which AI our clients use. DealCockpit™ can be accessed via Claude, ChatGPT, Cursor, n8n, an in-house agent or a custom script. Clients choose their assistant just as they choose the colleague they work with every day: the one they trust.
We provide the tools. The client retains control over the intelligence that orchestrates them.
Our commitments
The human decides. Always.
AI suggests, humans confirm. For any sensitive action, such as deleting files, modifying groups or inviting users, explicit confirmation is required. No autopilot. No irreversible actions without a human in the loop: this is an architectural red line, not a temporary limitation.
Our clients' data remains their own.
DealCockpit™ never uses its clients' data to train an AI model. Neither our own nor that of a partner. Never.
Total transparency on what the AI is doing.
When an AI assistant performs an action in a DealCockpit™ data room, it calls upon named, documented and auditable tools. The user can see what is going to happen before it happens, and what happened afterwards.
Included in all pricing plans.
MCP access is available to all our customers, with no premium options and at no extra cost. Innovation of this kind should benefit everyone.
What we will not do
We will not provide strategic recommendations nor opinions on the merits of a course of action. The quality of an analysis depends on the model that produces it, and we cannot vouch for its reliability. For strategic advice, our clients shall turn to a qualified human expert. This is in their best interests, and it is our firm belief.
We will not add AI features just to tick a box. When we offer an AI tool, it is because it delivers measurable value. Not because it allows us to write “powered by AI” in a sales presentation.
We see DealCockpit™ as a well-equipped platform, not as an autonomous agent. The more complex the transaction, the more valuable our traceability is. The higher the stakes, the more essential human oversight is.
Our role is to provide the clearest, most transparent and most traceable infrastructure possible, so that our clients can work faster, better, and with complete confidence. Not to sign off on analyses on their behalf.