Four triggers
Conversation created, conversation updated, conversation reopened and message created. Those four cover the whole cycle of a case.
The eventRules that assign, tag and reply with nobody watching. Macros that resolve a complete case in one click. And an AI that answers the simple things and prompts the agent on the rest, with the customer's history in front of it.
A rule is built in three blocks: what triggers it, which conditions must hold, and what Waya does. They combine with AND and OR, and run on every incoming message with nobody supervising them.
Conversation created, conversation updated, conversation reopened and message created. Those four cover the whole cycle of a case.
The eventChannel, team, agent, tag, country, browser language, source page, campaign and any custom attribute you have created.
The filterAssign to an agent or team, tag, reply, attach, mute, snooze, resolve, notify the team by email, send the transcript and fire a webhook.
The resultA macro chains together what an agent would do by hand: reply with the right template, tag, assign to the department and close. Saved replies are called up by typing a slash and the name, without leaving the text box.
Waya's AI works on two fronts. Facing the customer, it answers the questions already resolved in your help centre and in past conversations. Facing the agent, it drafts, improves, translates and finds the missing detail, without leaving the conversation.
It trains on your help centre, your past conversations and the FAQs you approve. It answers first, and when the case calls for it hands over to a person with the context already assembled.
It answers before anyone elseIt drafts the reply from the thread, improves a hastily written draft, translates it and finds the article that solves the case. The agent decides whether to use it.
It suggests, it does not sendIt finds what customers ask and your help centre does not yet answer, and proposes it as a new article. The knowledge base grows around what is actually being asked.
It spots the gapsIt keeps what matters about each customer between one conversation and the next: preferences, recurring problems, agreed terms. The next reply already knows.
It does not start from scratchIt does not invent the returns policy: it reads it from your help centre, from the PDFs you upload and from the pages of your own site you point it at.
You define how long the first reply, each subsequent reply and full resolution may take. Waya flags conversations approaching the limit before they breach it, and keeps a record of the ones that did.
It is for what was agreed in writing: corporate accounts, customers on a support contract, and teams reporting to a manager who asks for numbers.
Three ways to go looking for the conversation instead of waiting for it: the visitor who has been staring at your prices, the customer who needs telling something on WhatsApp, and the short notice that has to arrive even with no data.
It fires on what the visitor does: thirty seconds on the pricing page, a pass through the cart, a second visit that month. You pick the URL, wildcards included, and whether they are new or returning.
Automatic and continuousIt uses Meta-approved templates, with the fields that change in each message. You choose who receives it by filtering on a tag or a contact segment.
A one-off sendFor what absolutely has to arrive: a service outage, tomorrow's appointment, a due date. With a character and recipient counter before you send.
A one-off sendEvery inbox has its own hours, day by day. When a message arrives outside that range, Waya replies with what you wrote in advance, and the agreement clock does not start until the team is back.
In the demo we go through a week of your conversations and tell you which three rules to build first.