Skip to main content
Events are everything that happened, once. Every state change on the spine - an interaction recorded, an outcome written, a journey advanced, an entity merged - emits exactly one event you can consume. Events are how your systems react to conversations without polling, and how the outside world reaches back in: an inbound reply arrives as an event, and a running journey waiting on it wakes up. The alternative is what teams build today: cron jobs polling a call-log API, glue code diffing CRM records, a queue you run yourself to fan out “call ended” to three services. Events replace that with one contract - subscribe to what you care about, receive each change once, replay what you missed.

The primitives

On the spine

Events sit at the end of the spine and read all of it: every write to an entity, conversation, interaction, or outcome emits an event. They write nothing back to communication state - the stream is an append-only record of what already happened. In the other direction, events are the input side of Execution: an inbound message or reply lands as an event, and Flows & Journeys waiting on that signal resume.

What stays hidden

The broker. Which messaging system carries the stream, how partitions and ordering are maintained, how delivery is retried and deduplicated - none of that is a public surface. You see events with stable schemas and once-only semantics; the infrastructure that guarantees them is Unpod’s to run and swap without breaking your consumers.

Status

  • Production: the post-call webhook - every completed call delivers its transcript and structured outcome to your endpoint. Configure it per Hooks & Events; payload documented at post-call data.
  • In build: the unified event stream, signed webhooks, subscriptions, and replay - one feed across all channels and every spine primitive.

Go deeper

Hooks & Events

The event surface that exists today - call lifecycle hooks and the post-call webhook.

Post-Call Data

The payload the production webhook delivers - transcripts, summaries, extracted fields.

Outcomes

The structured records whose writes are the events you most often consume.

Flows & Journeys

The engine that reacts to events - outcome branching, waits, and inbound signals.