Automated support

An assistant that actually knows your business

Most chatbots repeat a script. This one checks your catalog before it speaks, so when it quotes a price, it is your price.
Why it does not make things up

How it differs from an ordinary chatbot

A raw AI model answers whatever sounds right. This one is forbidden from guessing the data that matters.

When someone asks "do you have the black jacket in medium?", the assistant does not answer from memory. It goes to your catalog, finds the product, checks the stock, and only then replies. If the product does not exist or stock is zero, it says so.

That is the whole difference between a useful assistant and a dangerous one. A bot that promises something you do not have does not save you work: it creates a problem with a customer.

What it can do

  • Answer prices, stock and availability by checking your catalog
  • Answer your FAQs: opening hours, shipping, payment methods, location
  • Build the order as the customer asks for things
  • Move the contact through your funnel when there is real evidence of interest
  • Hand over to a person when the situation calls for it

What it never does

  • Invent a price or a delivery time you never gave it
  • Confirm a payment: that is verified by a person or by the gateway
  • Promise stock that is not in the system
Vos escribís cómo habla. Le contás qué vendés, con qué tono tratás a tus clientes y qué no debe prometer. Es un formulario, no código. Y si algo no te gusta, lo cambiás y se aplica en el siguiente mensaje.
No black box

You are always in control

The assistant handles the conversation, but you see everything and can step in at any moment.
  • It switches off per conversation. One button in the chat and the assistant goes quiet for that customer. Everyone else keeps being served.
  • It never talks over a human. If someone on your team already replied, the assistant will not send an awkward second answer.
  • You see every message in your inbox, marked as sent by the assistant, with the same detail as if you had written it yourself.
  • You can try it before turning it on in a test panel, without any real customer seeing a thing.
Fair questions

Frequently asked questions

The ones we always get before people start.
Will my customers realise it is a bot? ▼
Probably, and that is fine. People today accept being answered by an assistant if the answer is fast and correct; what they will not forgive is waiting four hours for a price. You define the tone anyway, and you can have it introduce itself as an assistant from the first message.
What happens if a customer asks something unusual? ▼
It answers what it knows and, for what it does not, it says a human will reply and flags it in your inbox. It is the same judgement you would expect from a new employee: better to say "let me check" than to make something up.
Can I use the platform without the assistant? ▼
Yes. The inbox, the catalog, the quotes and the reports all work exactly the same with 100% human support. The assistant is one part, not a requirement.
Does it learn from my conversations? ▼
It uses the context of the ongoing conversation to stay coherent, and your catalog and FAQs as its source of truth. Whatever you teach it, you enter explicitly, so you always know where every answer comes from.

Try it with your own products

We load a few of your products and you message it as if you were a customer.