Using the Fallback topic

Important

Power Virtual Agents capabilities and features are now part of Microsoft Copilot Studio following significant investments in generative AI and enhanced integrations across Microsoft Copilot.

Some articles and screenshots may refer to Power Virtual Agents while we update documentation and training content.

The Fallback topic gets triggered when Microsoft Copilot Studio doesn't understand a user utterance and doesn’t have sufficient confidence to trigger any of the existing topics.

There are multiple ways to use the Fallback topic.

Offloading single question and answer pairs

If you have a large number of question and answer pairs for single-turn conversations, you can also consider offloading them to an external system and access them through the Fallback topic.

Microsoft Copilot Studio limits a chatbot to 1,000 topics. You can decide to manage exceptions to the Microsoft Copilot Studio topics that should generally do things for the end-user through the Fallback topic. With the Fallback topic, you can get answers for end-users that you retrieve from external systems using Power Automate, by calling a cloud flow from an action node.

Tip

Question answering in Azure Cognitive Service for Language provides cloud-based natural language processing (NLP) that allows you to create a natural conversational layer over your data. It is used to find the most appropriate answer for any input from your custom knowledge base of information.
Learn how to integrate question answering in Microsoft Copilot Studio.

Giving your chatbot a personality

Question answering in Azure Cognitive Service for Language provides a readily available "chitchat" data source, of various tones such as "professional", "friendly", "witty".

Chitchat lets your chatbot answer trivial questions, such As "how are you feeling today, bot?".

Integrate with generative AI models

You can also use the Fallback topic to call a generative AI model, such as those used by the Azure OpenAI Service, to handle more questions from your end users.

Through the system prompt and prompt engineering, you can also feed the model with instructions, rules and input data before it generates an answer.

Learning from the Fallback topic

You should track what your customers are saying that doesn’t trigger a standard Microsoft Copilot Studio topic. The phrases can enrich an existing topic’s trigger phrases, typically when the user utterance should have triggered a topic. The phrases can also lead to the creation of new topics to better serve your users as you understand better what they’re trying to accomplish.

Optimize trigger phrases and Natural Language Understanding