Temporal Explained • Determinism
Temporal
0:00 / 0:00
Temporal Explained • Determinism
5 763 просмотра · 4 г. назад
Temporal
17 тыс. подписчиков
5 763 просмотра · 4 г. назад
In this episode, Principal Curriculum Developer Tom Wheeler interviews Principal Engineer Dominik Tornow about the concept of determinism in Temporal Workflows. In Temporal, the Workflow Definition is deterministic orchestration code that must produce the same output given the same input, so non-deterministic work like LLM calls, API requests, and database queries belongs in Activities, which run outside the replay path and are retried reliably. This determinism is what makes Temporal well-suited for AI applications, since it can persist each step to recover from crashes, retry failed LLM calls, and resume long-running agent tasks without losing state.
Join the conversation at https://temporal.io/slack
Learn more
Use cases: https://temporal.io/use-cases
Developer resources
Docs: https://docs.temporal.io/
SDKs: https://docs.temporal.io/application-...
Support forum: https://community.temporal.io/