# Explore feature themes

> [!WARNING]
> **Preview:** This Pre-GA product is subject to the "Pre-GA Offerings Terms" in the [Service Specific Terms](https://cloud.google.com/terms/service-terms#1), offered "as is" with limited support. See the [launch stage
> descriptions](https://cloud.google.com/products#product-launch-stages) for more details."

Consider the following key feature aspects when creating adaptive agents to enable
universal design features.

## System or application control

Enable users to control the software system or app through the agent using their
preferred or necessary interaction modality---for example, voice commands
for users with visual impairments. Users should be able to provide feedback to
the agent, letting it personalize the interface, prompts, and interactions based
on individual preferences and context.

## Information seeking

Enable users to access information through the agent, potentially beyond
what's visible on the immediate interface. Users should be able to ask the agent
questions at any point, whether directly related to the main content, or
exploring broader or tangential topics. The agent should be able to then
retrieve information from the available content or from external sources such as
the internet, when possible, and should indicate whether the information source
is internal or external. The agent should be able to provide summaries and
answers to questions about the content, including broader topics.

## Content transformation

The agent should be able to take the outputs from the software system or app and
transform them to the user's preferred or necessary interaction modality, such
as visual content described audibly, or audio content captioned. The agent
should also work in tandem with the software system or app to deliver the
transformed content in a way that optimizes the user experience. For example,
when delivering audio descriptions of a video's visual content, the agent
shouldn't overlay these descriptions with the video's audio track but instead,
the agent should control the video playback and pause the video to insert the
audio description.

## Multimodal interaction

Emphasize the unique interactions that can come with a multimodal agent, which
can combine modalities for richer interaction. For example, when showing an
agent a picture of a restaurant's menu, the user might query the agent "what
time does this place open tomorrow". Even when this information is not on the
menu, the agent should be able to access that information elsewhere on the
internet and deliver the correct response.

## What's next

- [Key terms](https://developers.google.com/natively-adaptive-interfaces/guides/key-terms)
- [Case study: Explore feature themes](https://developers.google.com/natively-adaptive-interfaces/video/onboard#feature-themes)