How it works

I crawl your public pages once and keep the text as the agent's knowledge: product and service pages, policies, FAQ, shipping and returns, contact details. Nothing behind a login.

There is a ceiling on how much I load per site. A very large catalogue may need a search-based setup instead of a flat text base. I check that before I quote you, not after.

So the agent can answer "where is my order" with a real status, I connect it to your order and account systems. You expose the endpoints you are comfortable with, on your own server; I write the small piece that calls them. Every lookup is read-only and scoped to the customer asking.

This is the same whether you are on Shopify, headless, a custom stack or a subscription platform. If your fulfilment runs through a 3PL, I connect that too.

Two small pieces go on your side: one <script> tag on the pages where you want the widget, and a short proxy route on your own server, about a page of code, which I write for you. The proxy holds the key and tells the agent which customer is asking. The key never reaches the browser.

Before any customer sees it, your team tries it on a private link, against your real content and your real order data. It does not go live until you have said it is good, and the setup fee is not due until then either.

I check your site for changes as often as every 30 minutes. A full re-read only happens when a page you care about has actually changed, so the knowledge base moves when your site does. When you want the tone, the scope or a lookup changed, you tell me and I make the change. Your script tag never changes.

Once a month I send you the five things customers were most confused about, the objections turning up before checkout, and complaints as they surface. With each one, a specific change worth making. It is built from the real conversations, not a survey.