The text discusses the importance of standardizing event data in organizations, particularly in the context of an event-driven platform. It introduces three approaches to modeling events: the bare letter approach, which allows each process to define its own schema; the deep envelope approach, which pulls out commonalities into a top-level structure; and the shallow envelope approach, which uses a generic processor to handle multiple payloads. The text also explains how to materialize these models in the real world using an interface definition language (IDL) such as Protobuf, which is chosen for its flexibility, scalability, and compatibility with various programming languages. The article provides examples of how to define event models using Protobuf, including a temperature reading model that uses a deep envelope approach and an environment readings model that uses a shallow envelope approach. The text concludes by emphasizing the importance of selecting an appropriate approach for a given situation within the company and domain, as it affects risk reduction, maintainability, and generic processing.