Temporal

Why Temporal workflows need a frontend delivery layer

Temporal crash-proofs backend workflows. A durable sessions layer handles the rest: browser delivery, reconnect replay, and offline notifications.

Redis pub/sub limitations for Temporal frontend delivery

Redis pub/sub is fire-and-forget. Events published while a client is disconnected are gone permanently. Why it falls short for Temporal workflows and what to use instead.

Managing conversation history in a Temporal AI agent chat

Temporal's event history is an execution log, not a transcript store. Here's what breaks for long-running AI agents and how to handle history and delivery properly.

Delivering Temporal workflow output to multiple devices

Temporal workflows run once and produce output once. Getting that output to every device a user is watching from requires a fan-out and replay layer that Temporal doesn't provide.