A company sells all types of bicycles, bicycle parts, and accessories. You are creating a chatbot by using Microsoft Power Virtual Agent for the bicycle shop. When someone types in a bicycle brand…

Custom entities in Power Virtual Agents allow you to define lists of specific terms - such as bicycle brand names, 'helmet', and 'shoes' - so the bot can recognize and categorize them as accessory-related input. Once configured, the bot can use entity matches to trigger the correct topic routing, enabling automatic navigation to the accessories section.

The fallback topic is invoked only when no topic matches the user's input, making it unsuitable for proactively routing recognized accessory terms.

Smart matching is an automatic, non-configurable feature that handles fuzzy recognition of trigger phrases and cannot be directly set up to route based on entity terms.

Synonyms are a sub-feature within entity configuration used to broaden term recognition, but configuring synonyms alone does not provide the slot-filling mechanism needed to automatically route the conversation flow.

Slot filling enables the chatbot to automatically extract recognized entity values from the user's initial message and use them to satisfy required conversation variables, so the bot can immediately route to the accessories section without asking a follow-up question to clarify what the user wants.