An eager workflow in Flyte is a workflow defined with the @eager decorator that allows you to use native Python async/await syntax and constructs (like conditionals, loops, and exception handling) to compose tasks and subworkflows. Unlike static (@workflow) or dynamic (@dynamic) workflows, eager workflows materialize all task and subworkflow outputs as real Python values at runtime, enabling highly flexible and Pythonic execution graphs, but without compile-time type safety or static graph analysis. Eager workflows are executed as tasks under the hood and use the FlyteRemote object to launch tasks and subworkflows as soon as you await them, supporting true wall-clock parallelism on a Flyte cluster if tasks are independent and resources are available. See 
Flyte eager workflows documentation and 
Flyte blog.
Would you like more detail, such as code examples or limitations?
Sources:
- 
Flyte eager workflows documentation
- 
Flyte blog 👍 0 button 👎 0 button 💡 
Hint: Mention @runllm in the thread for followups.