An assistant that knows your price list by heart
The agent answers customers from your catalog, price list and policies — on your website and in messengers, around the clock. When a question goes beyond the knowledge base, it calls a person and hands over the whole conversation.
Price list, catalog, delivery and return terms, website pages, the best answers your operators have given. The agent answers only from these materials.
Website widget, Telegram, WhatsApp, VK. One set of answer logic, with a separate connector for each platform, because their APIs differ.
The agent does not stop at an answer; it suggests the next step: pick a product, book a demo or site visit, submit a request to the CRM.
A complaint, a custom discount, a question outside the base: the chat goes to an operator with full history, so the customer does not repeat themselves.
Questions the agent could not answer land in a report. We add the missing pieces, and next time it answers on its own.
One task, three different projects. Pick the one that looks like you and press “I want this” — your request arrives tagged, and we send an estimate in 1–2 days.
An online shop or service business where the owner or a single operator answers chats in between everything else.
What we build
Several channels, a catalog in your ERP, branches or pickup points, and customers asking about stock and order status, not just prices.
What we build
A chain or group with several brands and countries, where legal signs off on every customer-facing phrase.
What we build
Pilot from RUB 300,000. Beyond that we price by task and hours.
The agent answers from your knowledge base, not from the model’s general knowledge. If the answer is not there, it says so and hands the chat to a person. Discounts, returns and other sensitive topics can be limited to approved wording only.
Yes. We built ChatNeuron, a platform of AI agents for websites and messengers that answer from company documents and create requests in the CRM. It has 10+ paying customers. We bring that experience to your project, but the agent is built around your process and your systems.
Buttons work as long as the customer follows the menu. The agent understands free text, asks follow-up questions and brings the conversation to a request. You can still keep buttons where they are handier, for example to pick a branch.
In the cloud or on your servers — we decide at the start. If data cannot leave your perimeter under 152-FZ, we deploy the agent in a closed environment.
Name the process that eats time. You get a pilot range, not a 40-slide deck.