Dagger is best known as an open source container engine that vastly improves CI by allowing pipelines to be defined as code, in a team’s preferred programming language. But Dagger is actually a complete runtime and programming environment for distributed applications, with unique features such as repeatability, caching, tracing, platform independence, and a cross-language ecosystem of pluggable modules.
Besides complex build and test workflows, these features are perfect for complex AI workflows.
This talk discusses Dagger’s approach to managing the increasing complexity of AI workflows and demonstrates how it is a game-changer for building and running AI agents. It showcases how Dagger’s repeatability, caching and tracing features align perfectly with the demands of the exploding AI ecosystem.
Key Takeaways