Glossary
Rule-based and generative assistant – what does it mean?
Definition
Rule-based and generative assistant: a rule-based assistant follows a decision tree; a generative assistant formulates a new answer from the content.
The rule-based one is predictable but rigid; the generative one is flexible but can make mistakes without sources and handoff.
How it works
The assistant can answer customers in two ways. In the first mode, it follows pre-set branches: each question has one pre-written answer, and it cannot say anything else.
In the second mode, the system puts together an answer from the company’s own content, such as the price list and the website texts, and tailors it to the question. The two approaches can also work together in one assistant.
An example from a company
A visitor to an online store asks how much a product costs and when it will arrive. The assistant gives a precise answer on the price from the rule-based part, and formulates the answer on delivery from the content. If there is no content for the question, the assistant asks for contact details and hands the question off to a colleague.
What this means for business owners
Ask the provider where it draws the line between the two methods, and what happens if there is no answer in the knowledge base. The right answer: the assistant hands off and does not guess. If the line is not clear, your company may give its customers inaccurate or made-up answers, and correcting the error will be your work.
How we do it
Our assistant answers from the company’s own knowledge base, and the source is shown next to the answer. If there is no content, it does not make up an answer: it asks for contact details and hands the question off to a colleague. The Chat AI Starter monthly fee is €14.90 / month excl. VAT, with an allowance of 500 answered questions per month.
What to do now?
Do you have a question about your own website or online store? Request a demo or see pricing.