Learn conversation design.
How to decide what an AI agent says, get everyone to agree on it, and hand engineering a design they can build from. Written by Arian Fetahaj, founder of Klate.
What is conversation design for AI agents?
Somebody has to decide what the agent says when the refund is declined. That job is conversation design for AI agents, and it starts before any prompt.
Fundamentals · 9 min read
Read the definition
More on conversation design for agents.
How to design an AI agent conversation, step by step
Eight steps from a vague “we need an agent” to a design engineering can build from, each with the mistake that sinks it and the test for done.
Read the steps
A conversation design template that survives contact with engineering
Every section a design needs: turns, branches, tool calls, escalation and sign-off, each with a filled-in fragment from a card-dispute flow.
Read
Gathering requirements for an AI agent: a guide for business analysts
You were told to look into AI agents. Pin down seven answers before anyone builds, or engineering will invent them and you will find out in UAT.
Read
Agent flow documentation: what to write down before you build
Escalation rules belong in the design; retry timers belong in the platform. Draw that line and the design stays the record of what was agreed.
Read
Conversation design vs. prompt engineering
Conversation design decides what the agent should do; prompt engineering gets a model to do it. Run them out of order and you tune the wrong behaviour.
Read
The happy path is not a design
Timeouts, angry customers, out-of-scope asks, users probing the rules. The four branches nobody demos, and the first ones a real customer finds.
Read
Getting a client to sign off on an agent design (before you build it)
How to turn a nodding workshop room into written approval of one versioned design, so month four never opens with “that’s not what we agreed”.
Read