The Prompt-to-Product World Just Got One Step Closer

This newsletter is part of The Assetizer, GenTwo's thought leadership platform.


 

This week GenTwo released its GenTwo Connector for Claude. It uses the widely adopted Model Context Protocol (MCP) to connect users of the GenTwo Pro platform with their Claude agents. 

While this is neat, I’ll be the first to admit that in and of itself the connector is not groundbreaking news. MCP is being used to build such connections all the time.

But put it in the broader assetization context, and I think it’s definitely worth a closer look. 

As Assetizer readers will know, we believe the end state for the investment industry infrastructure is radical automation: a “prompt-to-product” world. Our connector is an excellent example of a first step. 

To get the bigger picture I had a chat with Ursin Brunner, the head of AI at GenTwo, about the connector and also the wider context of where the agents are leading us. 

The Once and Future Agent

Before we get into the details, there are two pieces of context you need to know. 

GenTwo Pro is the platform GenTwo’s clients use to create, structure, launch and administer financial products. It coordinates the work of building a product, and then, once the product is live, handles the running of it: valuations, orders and rebalancings, coupons and payments, approvals, and the rest. The first Swiss B2B platform of its kind, GenTwo Pro can be seen as an initial move in our mission to streamline the securitization process.

But it was always a beginning. For us, the end state is a world where you describe the product you want, and structuring, documentation, compliance and launch happen without you touching a form. We even made a short film about it two years ago, in which a gallery owner turns a bronze sculpture into an investible product in under thirty seconds. The connector is the first piece of that film to arrive in the real world.

Walk Before You Run

Admittedly, we are still far from full automation. What GenTwo has released, as Ursin told me, is “really a new channel. Instead of a web interface for humans, it’s the interface for agents.”

With the connector, existing GenTwo clients who use Claude can basically chat with their GenTwo-launched products and with our platform. “You want to see your portfolio, how products change over time, dive into the term sheets, see which tasks are pending, etc. All this and more you can do.”

What you can’t do, yet, is change anything. “It’s read-only for now, partly to get a feeling for what users actually want to do,” Ursin said.

That is set to change. “Everything you can do on the platform today, and some things you still do by email, mid- to long-term can be done by the client’s AI agent through this connector. It’s going to be a fully-fledged channel. Read-only is just the start.”

Nor will it be limited to Claude. Ursin says the company is planning to release the ChatGPT version this quarter, and eventually will make it available for all AIs.

From Prompt to Product

What’s interesting here is not necessarily what this first connector can do now, but what it says in terms of where things are going. Ursin agrees. 

“In the future a large part of the economy is not going to be humans going to a website and doing something,” he told me. “It will be agents acting for different players, accessing other businesses directly.”

We can think of MCP as an API equivalent, he added, one built specifically for agents and designed to be universal. 

That means increasing automation. “Right now, it’s usually still a human using an AI and then doing something. But we see a future where agents do things directly with our platform, maybe without humans in the loop.”

Prompt-to-product has come one step closer. 

Tom Lyons, GenTwo

For more:

• AI in Finance: The Brainy Black Box (my podcast with Ursin)

• Rethinking Securitization from the Ground Up (my podcast with Patrick Loepfe, Chairman and Founder of GenTwo, on his vision for securitization and the industry; includes a discussion of the coming wave of automation)

• Wave 3: The Radical Automation of the Investment Industry