Okay, so what can we put in place then to accelerate agent development but also reduce our risk so we can get kind of a win-win?

Well, you're setting a high bar, and it's totally possible.

One of the best things you can do to unlock developers and agent coding is to establish a catalog of shared services and resources, which exists in two parts.

The first provides a shared platform that centralizes AI model registries, knowledge sources such as MCP servers, APIs, and agents, making them easy to discover and use, along with shared identity and access management, policy engines, and reporting for consistency and observability.

We call this the governance hub, and it's directly tied to Agent 365 and architecturally part of what we call the platform landing zone.

The second part, which we call the intelligent data platform, consists of the shared data components themselves, including core operational data, as well as the underlying data that agents connect to via MCP servers and APIs.

These are managed data sources spanning structured and unstructured data where each can be leveraged as tools for agents.

Is there an accelerator then to deploy AI landing zones?