Flows are sequences of steps and actions that define how your AI Assistant interacts with users. They empower your AI assistant to go beyond answering questions, enabling it to take action and get work done.
Flows are currently triggered by intents which represent the users purpose or goal, e.g.: "where is my order?". Unlike traditional rule-based chatbots, Frontline's flows can be triggered via intents at any point of the conversation, providing a more natural and fluid experience to end users.
For example, you can build a simple flow to track an order status that is triggered by an intent, e.g.: "where is my order" and routes to a sub-assistant that has specific instructions to retrieve the customer's order number. The sub-assistant can leverage the knowledge base to help the customer locate their number in case they are unable to find it (providing a much better experience), and once the number is provided it can execute an API call retrieve the order status, parse the response and provide an answer just like a human support agent would.
Creating a flow
To create a flow head to assistants > flows and access the flows sub-section and click on "Create Flow". Choose a name and description (optional), these won't affect the flows behavior but will help your team identify it's purpose or goal.
Editing a flow
To edit a flow head to assistants > flows, access the flows sub-section and click on the desired flow.
Deleting a flow
To delete a flow head to assistants > flows, access the flows sub-section and click on the three dots. Once the dropdown menu opens, click on delete and confirm it's deletion. Take into account that deleting a flow can impact the assistants behavior.
Accessing the designer
The designer is where you can build and customize your flow, it's similar to a canvas. To access the designer head to assistants > flows, access the flows sub-section and click on "Go to Designer".
Changing the Status of a Flow
Flows can be set to Live, Offline or Draft. If a flow is live, the AI Assistant will be actively listening for intent matches on every message and route conversations to that specific flow if a match occurs. If the status is set to Offline, the AI Assistant won't be listening for intent matches, and the flow won't be executed. If the status is set to draft, the AI Assistant will be only listening for intent matches on test conversations coming from within the Frontline dashboard.