How it works
How AI chatbots actually work
From the visitor's question to the answer on screen: retrieval, context, the language model and why grounding in your content matters.
In short
An AI chatbot takes the visitor's question, finds the most relevant passages from your indexed website content, and asks a language model to answer using only those passages.
The model supplies the language; your website supplies the facts. That split is why a grounded chatbot can be trusted with prices and policies.
Key facts
Key facts
- The pattern is called retrieval-augmented generation, or RAG.
- Your pages are stored and searched at question time, not baked into the model.
- The model is instructed to answer only from the supplied passages.
- Answers stream word by word, which is why they appear as they are written.
Step by step
- · The visitor types a question into the widget.
- · The system searches your indexed content for relevant passages.
- · Those passages, plus your instructions and the conversation so far, go to the model.
- · The model writes an answer grounded in that material.
- · The answer streams back into the widget.
Why grounding matters
A language model without your content will happily invent a price. With your content attached and a clear instruction to use only it, the same model becomes a careful reader of your website. The difference between a chatbot you trust and one you switch off is almost entirely this.
What the bot does not have
It has no access to your CRM, order database or calendar unless you connect one. It knows what you published and what you added as extra knowledge, and it should say so when asked something beyond that.
The product behind this guide
A chatbot trained on your own website, live today
Defaltam builds AI chatbots that read your public pages and answer with your real prices, services and policies. Support is €79/month, Sales is €149/month, both start with a 5-day free trial, and installation is one script tag.
FAQ
Questions people ask about this
How does an AI chatbot work?
It searches your indexed website content for passages relevant to the visitor's question, then asks a language model to write an answer using only those passages. The model supplies the wording; your site supplies the facts.
Is the model trained on my data?
No. Your content is stored and retrieved at question time. Nothing is baked into the model itself, which is why re-indexing updates answers immediately.
What is RAG?
Retrieval-augmented generation: the technique of retrieving relevant documents and passing them to a language model so its answer is grounded in real source material.
Can it access my order system?
Not by default. It answers from published content unless you build an integration.
Related
Keep reading.
How to stop a chatbot from making things up →
Hallucinated prices and invented policies destroy trust. The four controls that keep a website chatbot honest, and how to test for them.
Running a multilingual AI chatbot →
How a modern chatbot answers in the visitor's language from a single-language website, where it goes wrong, and what to translate anyway.
Website chatbots and GDPR: what you need in place →
What data a website chatbot collects, what your privacy policy must say, and how to handle transcripts and consent under GDPR.
Designing a chatbot handoff to a human →
When a bot should stop answering, what it should collect before passing over, and how to make the handoff feel like service rather than failure.
Website chatbot security basics →
Prompt injection, data leakage, abuse and rate limits — the security questions to ask before putting an AI chatbot on a public website.
Why most website chatbots fail →
Chatbots fail for four predictable reasons: no clear job, thin content, no handoff and nobody reading the logs. Each has a straightforward fix.
AI chatbot plans and pricing →
Support €79/mo, Sales €149/mo, 5-day free trial.
How to add an AI chatbot to your website →
The ten-minute install guide, platform by platform.