Website Assistant Over Your Own Content
Answers only from your pages and documents, in both languages, with lead capture at the end and a human hand-off when it does not know.
Services
AI Assistants & RAG Systems
Distributors drowning in quotation requests, clinics with a WhatsApp inbox nobody can keep up with, and any operator with documents that need reading and a CRM that needs feeding.
An assistant is only as good as the data behind it and the checks around it. We build assistants over your own catalogue, content and records, on WhatsApp and the web, that answer in the language the customer wrote in, hand off to a person with the conversation attached, and never invent a price, a spec or a diagnosis. Behind them, retrieval and extraction pipelines that turn documents into structured data your systems can use.
Who this is for
Distributors drowning in quotation requests, clinics with a WhatsApp inbox nobody can keep up with, and any operator with documents that need reading and a CRM that needs feeding.
What we deliver
- Website and WhatsApp assistants over your own content, with dialect test suites and a human hand-off that pauses the AI
- Sales assistants that qualify an enquiry, draft the quote from your catalogue and write to the CRM, with a person approving every send
- Document and record extraction: LLM structuring against a schema, entity annotation, retrieval over your files, confidence scores on every field
- The five checks on every system: the model never decides, data stripped before it leaves, schema-validated output, an audit trail, both languages including mixed
AI Assistants & RAG Systems
WhatsApp Assistant With One Integration
On your verified number through a provider; booking system, Salla, Zid, WooCommerce, HubSpot or Zoho as the first integration; dialect test suites.
Sales Assistant Wired to the CRM
Extracts the request, matches it to the catalogue with confidence scores, drafts the quote from your price list, writes to the CRM. A person approves every send.
Document Extraction With LLM, RAG and NLP
Structuring against a schema, entity annotation, retrieval over your own files, a confidence score on every field, and the gaps shown rather than filled.
The Five Checks
The model never decides; data is stripped before it leaves; output is validated against a schema; every step leaves a trail; both languages, including mixed.
Related reading
A WhatsApp Assistant for a Clinic: The Build Checklist for Booking, Boundaries, PDPL and the Human Hand-off
A clinic assistant should do four things well: book, remind, answer the practical questions, and hand the rest to a person with the conversation attached. Everything clinical is the rest. Here is the checklist we build against.
Read More
From a WhatsApp Request to a Priced Quote: An AI Quotation Assistant for Building-Materials Distributors
Most requests for quotation arrive as a WhatsApp message, a voice note or a photographed spec sheet. An assistant can read them, match them to your catalogue and draft the quote. What it must never do is set the price or promise stock.
Read More
Spec Sheets Into Structured Product Data: Making a Building-Systems Catalogue Readable by Engineers, Search Engines and AI Assistants
A product page that is a photo and a link to the manufacturer's PDF is invisible to all three readers who matter: the engineer, the crawler and the assistant. One structured record serves all of them.
Read MoreFAQ
Which platforms and integrations?
WhatsApp Business API through a business solution provider; Salla, Zid, WooCommerce, HubSpot or Zoho as the first integration; your ERP or price list through rules the model never touches.
Where does our data go?
Identifiers are stripped before any text reaches a language model, the application and its logs are hosted in the jurisdiction your law expects, and the model call is documented as a transfer with a written risk assessment. Every bot discloses it is an assistant and offers a person.
What does it cost?
A website assistant from about USD 2,400; a WhatsApp assistant with one integration from about USD 4,800; a sales assistant with CRM from about USD 8,500. Each ships with a maintenance plan of USD 250 to 900 a month, because the platform, the catalogue and the model all change.
Describe one process. We map it before we quote.
Send one week of real requests or messages, names removed. You get a written map of what an assistant would answer, what it would hand off, and a fixed price.
Prefer to talk? Message us on WhatsApp