Conversation Graph
A conversation graph models communication as a connected structure of messages, threads, entities and people — instead of a flat list of channels — so any two related things can be linked, queried and traced.
Flat-channel chat treats every conversation as an isolated stream: a channel has messages, a thread has replies, and there's no first-class way to say "this discussion is about that invoice" beyond pasting a link. A conversation graph makes those relationships part of the data model — messages connect to threads, threads connect to entities (a task, a risk, a customer), and people and agents connect to all of it as participants.
The practical payoff is that a graph is queryable in ways a flat channel list isn't: you can ask "everything about this project," "everything this person decided," or "everything an agent did this week" and get a structured answer instead of a manual search through scrollback.
It also changes what search and memory can do. Once conversation is a graph rather than a pile of channels, an org-memory layer can build entity profiles from it, and permission-trimmed search can reason about what's connected to what, not just which words match.
Conversation Graph, in the product
Aanty's kernel is an event-sourced conversation graph: typed messages, entity-linked threads, and principals (humans, agents, integrations) are all nodes the graph connects. That structure is what makes the Briefing, the decision ledger, org memory and permission-trimmed search possible on top of the same data, instead of four bolted-on features.
Related terms and pages
See Conversation Graph in a real workspace
Bring a channel from wherever your team works today. In fifteen minutes we'll show what conversation graph looks like on a real conversation, not a slide.