13.2Asking a question

Ask Bmator searches the workspace you select, sends relevant passages to an AI model, and returns an answer based on your documents. Your conversations are saved and searchable, and follow-up questions keep the context of the conversation.

Every answer costs a small amount of AI credit, which is debited from the workspace you ask in. If the workspace balance cannot cover the required amount, your question is refused rather than half-answered.

Starting a conversation

1. Open Ask Bmator from the left menu. Alternatively, select Ask AI in the top bar for a quick question, where a link lets you continue the conversation in Ask Bmator.
2. Choose which workspace provides the context and covers the cost by setting Ask in workspace if you have access to several workspaces. This setting determines which workspace's knowledge answers the question and receives anything you save, without changing the active workspace chosen in the top bar.
3. Type your question in the text box.
4. Optionally, select Attach files to this question to include files with your inquiry. You can choose to keep these files in your Library.
5. Select Model if you want to pick a specific AI model for this conversation. If you do not choose one, your default model is used.
6. Adjust Answer settings if you need to configure parameters such as search breadth or cost bounds.
7. Submit your question to receive an answer citing sources from your material.

Asking within a project

Selecting Ask about this project on a project's Overview page opens Ask focused on that project alone. The answer consults only that project's documents and its own records, including its work, dates, and people.

While scoped to a project, a line appears naming the project alongside Ask about everything instead, which lets you return to a broader scope. In this view, selecting Save as work item adds the answer directly to the project.