In today's dynamic technological landscape, AI-driven applications are gaining increasing significance. To address the challenges and complexities of these applications, robust and flexible architectural approaches are essential. In this talk, we will explore why Event Sourcing and Domain-Driven Design (DDD) are excellent methodologies for developing applications based on language models (S/LLMs).

Event Sourcing allows the complete state of a system to be represented through an immutable sequence of events. This not only provides precise traceability and auditability but also the ability to easily restore past states of the system. These characteristics are particularly valuable in AI-driven applications, where the traceability of decisions and the ability to optimize and realign models are crucial.

Domain-Driven Design (DDD) promotes a deep understanding of the domain and a clear separation of responsibilities within a system. By applying DDD principles, developers can ensure that their AI models are closely aligned with business requirements and processes. This leads to more accurate and reliable models that meet the actual needs of users.

In this talk, we will present practical examples and best practices of how Event Sourcing and DDD can be leveraged to facilitate and enhance the development and maintenance of LLM-driven applications. We will also address challenges and discuss solutions to support the integration of these methodologies into existing systems.

This talk is aimed at developers and architects interested in implementing AI-driven applications. Basic knowledge of Event Sourcing and Domain-Driven Design is assumed. The talk includes technical details and practical examples to illustrate the discussed concepts.