A palette of steps, a canvas with the Start step already placed, and a property panel for whatever you select. Type variables into any message, walk the flow in the test chat beside the canvas, and deploy once validation passes.
No code ✦ Draft until you deploy ✦ Partner API for every part of a journey
The most guest-obsessed brands trust Lodgestory to power 50M+ conversations a year
System variables are filled by the engine and read-only: the customer's name, phone, channel and latest input. Local variables live for one run and hold what the journey captured, an arrival date, a party size, an intent. Campaign variables are organisation-wide, so one journey can read what another, or a campaign, wrote. Type {{scope.name}} into any message and the value lands in the text; a missing variable leaves the placeholder visible instead of failing.
Drag from a step's output handle to the next step's input. Labelled links such as success, timeout and fallback decide the path.
Every step's fields, limits and previews live in the property panel. What you see is what the channel sends.
The test chat runs the journey beside the canvas without touching a customer, and previews what tickets and stage changes would do.
Deploy checks the structure and every Time Condition, points at the exact field when something is off, and flips Draft to Live.
Open Test and a chat panel appears beside the canvas. Type as the customer; each step answers in the sandbox. Steps that would write something show a one-line preview instead: "Ticket would be created → workflow / category / priority". A Time Condition asks which clock to test on and fast-forwards a hold. When it reads right, Deploy validates and goes live; moving back to Draft ends running sessions on purpose.
A greeting, a question and a hand-off take five minutes. The test chat tells you if it reads right.
About the builder itself.
A palette of step types on the left, the canvas in the middle with the Start step already placed, and a property panel on the right for the selected step. The toolbar saves, opens the test chat and deploys. The Variables panel lists system, local and campaign variables, and the AI Agent step opens its own pane with Overview, Persona, Model and knowledge, Tools, Exits and variables, and Conversation sections.
System variables are read-only and filled by the engine: customer name, phone, channel, latest input. Local variables live for one run and store what the journey captured. Campaign variables are organisation-wide and shared across journeys. Any message can interpolate {{scope.name}}; a missing variable leaves the placeholder visible rather than erroring.
It runs the journey in a sandbox beside the canvas. You type as the customer and watch each step answer. Ticket and workflow-stage steps show a one-line preview instead of writing anything, a Chat Session step lets you pick the outcome, and a Time Condition can be tested as if it were any day and time, fast-forwarding holds.
That there is a Start step plus at least one other, that every link points to an existing step, that each Time Condition has a valid time zone, uniquely named slots with rules and no end time equal to its start, and that a Create Ticket step's workflow is enabled on the channel. A failure points at the exact field. Success flips the journey from Draft to Live.