Using Wago
AI, knowledge & catalogues
Teaching Wago your facts and products, and the steps that answer from them.
Wago's AI steps answer from what you teach it: your knowledge entries and your catalogue. Teach it once in Settings, then use it from any workflow.
Knowledge
Settings → Knowledges holds the facts a customer might ask about: opening hours, delivery areas, returns, how to book. Each entry has a title, the content and tags. Short entries that each answer one thing work best.
Catalogue
Settings → Catalogues holds your products: a name, a description, the price and currency, a SKU and a picture.
The steps that use them
| Step | What it does |
|---|---|
| Search knowledge | Finds the passages in your knowledge that answer something, and keeps them in a variable for an AI step below. |
| AI reply | Lets AI answer in your words — optionally from your knowledge — and sends that answer as the reply. |
| AI answer | Works something out with AI from a prompt you write and keeps it in a variable. Nothing is sent; later steps use it. |
| Answer from catalogue | Works out which product the customer means — even in their own words — and replies with its picture
and a caption of name, price and description, or plain text when there is no picture. It branches found / not found, and fills {item.name}, {item.price} and friends for later steps. |
| Send products | Sends items from a Meta Commerce Manager catalogue as WhatsApp product messages. Needs that catalogue and its SKUs set up at Meta. |
How AI is counted
AI is measured in tokens — roughly, pieces of words. Input tokens are what the model reads (your knowledge, the question, the instructions); output tokens are what it writes. Every plan includes an allowance of each, and before an AI step runs Wago checks there is room for it. Past the allowance, tokens come out of top-ups. See Plans & billing.