Integrate Multimodal Component into Notebooks Environment

[agusakow]

We need to integrate the Multimodal component from issue 162 into the Notebooks environment.

This will bring the following benefits:

• UI Flexibility - users can switch between Plot, Table, and Text representations in the UI.
• Improved latency - specific answers are only requested/generated when the user switches to the corresponding tab

Technical Requirements:

• Use a unified Python API for the multimodal component. It must automatically detect the Jupyter environment and handle the Jupyter widget render or use one shot server for terminal usages.
• OnDemand Fetching - request specific answer types from the kernel only upon tab switch events.