The AI Assistant is available across Productive and can use context from whatever you're currently viewing (like an open task or project).
You can refine requests or ask the Assistant to regenerate outputs until you get the results you need.
In this article, you’ll learn how to access the Assistant, navigate the interface, and understand how context and references shape its responses.
📌 Important: AI-generated responses may be inaccurate or incomplete. Always review results before applying changes or relying on outputs.
Accessing the AI Assistant
You can open the AI Assistant from the main navigation.
Click the AI Assistant icon in the upper left to open it
To start a new conversation, simply start typing or click New chat at the top of the Assistant to start fresh
To view previous conversations, select a chat from the chat history list
Drag the sidebar left or right to resize the Assistant interface
Click the collapse icon to close the Assistant sidebar
📌 Note: Search in chat history is case-sensitive.
Check out the interactive walkthrough below to see where the AI Assistant appears in the interface and how to use these controls.
What You See in the Assistant Chat
When you open the AI Assistant, you’ll see:
A chat input for natural language prompts.
Suggested prompts to help you get started.
Tip: If you’re not sure what to ask the Assistant, these prompts are a good way to explore what it can do.
Options to add attachments or reference content.
Reminder: The AI Assistant is iterative. You can refine your questions, follow up on responses, or have it redo outputs to better match what you want to see.
How the AI Assistant Uses Context
The AI Assistant interprets requests based on context. Context can come from:
The item currently open (task, project, report, etc.)
Items explicitly mentioned in the message
Items manually added using referencing
If no item is referenced explicitly, the Assistant relies on the current context to decide what the request applies to.
Specifying Which Items to Act On
You can reference specific items when interacting with the AI Assistant. This helps ensure actions are applied to the correct item.
In general, how references are resolved depends on:
Whether an item is currently in focus
How clearly you specify the target in the message
Referencing Examples (click to expand)
Referencing Examples (click to expand)
1) Implicit task references (default behavior)
If a task is currently open and you do not explicitly mention another task, the Assistant assumes the request applies to the open task.
📌 Example scenario
You previously created two tasks: Tuesday and Wednesday
You open a third task: Friday
The chat shows that this task is the current context
User input:
Set me as assignee
Expected behavior:
The assignee is set on Friday.
2) Explicit task references
If you explicitly name a task, the Assistant must follow that reference, regardless of the current context.
📌 Examples
Set me as assignee on the Tuesday task
→ Assignee is set on TuesdaySet me as assignee on the Wednesday task
→ Assignee is set on Wednesday
3) Unclear or ambiguous references
If a request cannot be mapped to a specific task with certainty, the Assistant does not make assumptions.
📌 Example:
Set me as assignee on the previously created task
Expected behavior:
The Assistant will ask you to clarify which task you mean. Only after you provide a clear reference does the Assistant perform the action.
Tip: You usually don't need to specify items manually. The AI uses context automatically. Only add explicit references when working with multiple items or when you want to be extra clear.