If this, then that
Trigger on a keyword, a channel, a customer attribute, an SLA breach or an inactivity timer, then assign, tag, reply, escalate or close.
FeatureAutomations
The routing and chasing should not be someone's job.
Rules that assign, tag, escalate and follow up on their own, plus chatbot flows for the paths you want handled exactly the same way every time.
No code requiredNo credit card requiredLive in under 5 minutes
Rules for the work that is the same every time.
Triage, tagging, chasing and closing are predictable, repetitive, and exactly the sort of thing a person should not spend a morning on.
Routing by topic and team
Send billing to finance and integrations to engineering, with a fallback so nothing lands in nobody's queue.
Chatbot flows
Build the paths you want handled identically every time, such as collecting an order number before anyone picks it up.
Follow-ups that fire
Chase a customer who went quiet, close a thread after a set period, or nudge a teammate before an SLA breaches.
Reach your own systems
Call a webhook or your own API as a step, so a rule can create the ticket, refund the order, or update the CRM.
Reversible one at a time
Each rule carries its own run count and log, and switches off on its own. A rule doing nothing, or the wrong thing, is easy to spot and easy to stop.
Building
Built by the people who answer the tickets.
The person who knows the routing rules is the support lead, not an engineer. The builder is visual, so writing a rule does not need a sprint.
- Visual builder, no code and no deploy
- Conditions on channel, keyword, customer or SLA
- Branches for the cases that need different handling
- Webhooks when a rule does need to reach your systems
In the queue
You can see what fired and why.
An automation you cannot audit is one you will eventually turn off. Every rule that touches a conversation leaves its name on the thread.
- Each rule logs what it did, on the conversation
- Turn one off without unpicking the others
- Counts per rule so dead ones are obvious
- Manual override on any thread, at any point
Three steps to a quieter queue.
Start with the rule you would write first if you had the time.
Pick the repetitive one
Usually routing by topic, or chasing customers who went quiet. One rule is enough to feel the difference.
RoutingFollow-upTagging
Build it
Set the conditions and the actions in the visual builder. Start with a narrow condition so the first version touches few conversations.
ConditionsActionsBranches
Turn it on and watch
Every rule logs what it touched, so you can tell within a day whether it is helping or needs a condition tightening.
LogsCountsOverride
No codeBuilt by support, not engineering
$3Per extra agent, once the plan's seats are used
$0Free plan, no card, no expiry
5 minFrom signup to a live rule
What teams ask about automations.
Do I need an engineer to set these up?
No. The builder is visual and the conditions are the ones a support lead already thinks in: channel, keyword, customer attribute, SLA state. Webhooks are there when a rule genuinely needs to reach your own systems.
Can I see what a rule did?
Yes. Every rule that touches a conversation logs its name and action on the thread, and each rule carries a count so the ones doing nothing are easy to spot.
What if a rule gets it wrong?
Any teammate can override on the thread, and you can turn a single rule off without touching the others. Start with a narrow condition and widen it once the log shows the rule doing what you expected.
How is this different from the AI agent?
Automations are deterministic: the same input always produces the same action, which is what you want for routing and SLA work. The agent handles the open-ended questions where the answer is not a rule.
Can we host it ourselves?
No. Chattypie is hosted by us.
Stop triaging by hand.
Write the rule once.
Free plan, no code, and your first routing rule live in about five minutes.