AI Flows
AI Flows are multi-step conversation workflows for the Chat Bot tools inside Stream Studio. Combine triggers, conditions, static replies, and AI generated responses in a visual editor.
Flow Features
Visual Flow Canvas
Build flows on a drag and drop canvas. Connect nodes to create branching conversation paths that AI follows when triggered.
Entry Conditions
Start flows with AND or OR condition groups. Use conditions such as message content keywords to decide when a flow activates.
Conditional Branches
Add check nodes that branch a flow based on conditions such as user level (subscriber vs regular viewer) and route to different responses.
Static and AI Responses
Each response node can be a static text reply or an AI generated response with a context hint. Use variables like {user} and {message} in either type.
Test and History
Test individual flows or every flow at once from the Flows tab. Review recent execution history per flow to see how AI handled real chat messages.
Priority and Cooldowns
Assign a priority (0 to 100) to control which flow runs first when multiple flows match. Set a per flow cooldown (0 to 3,600 seconds) to limit repetition.
Node Types
Entry Condition Group
The starting node of every flow. Define one or more conditions joined by AND or OR logic. The most common condition is a message content keyword.
Check Condition
A branching node that evaluates a single condition and routes the flow to either a true or false output. Use it to handle subscribers and regular viewers differently, for example.
Static Response
A response node that sends a fixed reply. Supports template variables such as {user} and {message}. Use for predictable, scripted replies.
AI Response
A response node where AI generates the reply. Provide a context hint to guide what AI says; the personality and conversation memory settings still apply.