Cloudscape Design System
  • Get started
  • Foundation
  • Components
  • Patterns
  • Demos
  • GitHub 
Cloudscape Design System
  • Get started
  • Foundation
  • Components
  • Patterns
  • Demos
  • GitHub 
  • About

Patterns


  • General

    • Actions
      • Global actions
      • In-context actions
    • Announcing new features
    • Announcing beta and preview features
    • Communicating unsaved changes
    • Data visualization
      • Chart metric drill down
    • Density settings
    • Disabled and read-only states
    • Drag-and-drop
    • Errors
      • Error messages
      • Validation
    • Empty states
    • Feedback mechanisms
    • Filtering patterns
      • Saved filter sets
      • Filter persistence in collection views
    • Hero header
    • Help system
    • Image magnifier
    • Loading and refreshing
    • Onboarding
      • Hands-on tutorials
    • Selection in forms
    • Service navigation
      • Side navigation
      • Top navigation
    • Service Dashboard
      • Static dashboard
      • Configurable dashboard
      • Dashboard items
    • Secondary panels
    • Timestamps
    • User feedback
    • Variables

  • Generative AI patterns

    • Pattern abstraction
    • Generative AI chat
    • Ingress
    • Generative AI output label
    • Generative AI loading states
    • Progressive steps
    • Follow-up questions

  • Resource management

    • Create resource
      • Multipage create
      • Single page create
      • Sub resource create
    • Delete patterns
      • One-click delete
      • Delete with simple confirmation
      • Delete with additional confirmation
    • Edit resource
      • Page edit
      • Inline edit
      • Attribute editing
    • View resources
      • Table view
      • Card view
      • Split view
      • Table with expandable rows
    • Resource details
      • Details page
      • Details page with tabs
      • Details page as a hub
  1. Cloudscape Design System
    • Patterns
    • Generative AI patterns: Follow-up questions
    1. Cloudscape Design System
    2. Patterns
    3. Generative AI patterns: Follow-up questions

    Follow-up questions

    Ways to request additional information from users to help generate more accurate and relevant responses.

    Published: July 16, 2025

    On this page

    1. Key UX concepts
    2. Criteria table
    3. Common use cases
    4. General guidelines
    5. Writing guidelines
    6. Accessibility guidelines

    Did this page help you?

    Tell us more - optional

    1000 character(s) available. Do not disclose any personal, commercially sensitive, or confidential information.

    Key UX concepts

    AI systems display follow-up questions to collect more input from users to provide more precise responses. These questions can appear in various interfaces, from conversational chat experiences to structured forms.

    Lead with benefits

    Help users understand the value of additional context they provide. Highlight the benefits of enabling AI systems to generate better output via collection of more information to improve AI reasoning, and the expected outcome.

    Gather context at the right time

    Decide whether to gather information before showing output or during the interaction. For high-impact decisions or when confidence is low, collect details first. For other scenarios, start with an initial response and ask follow-up questions to refine results.

    Streamline user choices

    Provide clear options for users to proceed, skip, or exit at each step of the interaction. This ensures users feel in control of the experience and reduces friction.

    Guide users to necessary information

    Help users access information that isn't readily available. Provide links or instructions on where to find required data when appropriate. This ensures users can provide accurate information even when it's not immediately at hand.

    Match input mechanism to question type

    Display input methods that complement the type of information to be collected. For example, in chat experiences, open ended text responses and single selection mechanisms often work best. Questions with a finite set of answers, such as selecting resource IDs or timestamps, should use structured containers with a form . This reduces errors and improves data quality.

    Criteria table

    Information type

    Input method

    Examples

    Clarifying user intent

    Open ended text response using the chat prompt input.

    User needs to explain a use case. Multiple complex aspects need to be addressed in one response.

    Guided selection

    Present recommended inputs with support prompt group.

    There are up to five specific valid choices. You need to guide users to valid inputs. Questions change based on previous answers.

    Form-based input

    Present a form within the dialog box .

    Multiple specific data points are needed. Information has validation rules. Structured data is required.

    Progressive steps input

    Use form-based input within progressive steps.

    A specific execution step needs input to proceed. Details are needed for upcoming steps.

    Common use cases

    Clarifying user intent

    When users provide ambiguous or incomplete information, open-ended questions help gather details about their specific context and current situation.

    Generative AI assistant

    JD
    I need to optimize EC2 costs across multiple regions.

    Optimizing EC2 costs across multiple regions is an important task for managing cloud expenses. Here are strategies you can implement to optimize your EC2 costs:

    1. Use Reserved Instances (RIs)
      • Purchase RIs for predictable, long-term workloads
      • Take advantage of regional benefit and size flexibility
      • Consider Savings Plans for more flexibility across instance types and regions
    2. Implement Auto Scaling
      • Set up Auto Scaling groups to automatically adjust capacity based on demand
      • Use target tracking scaling policies to maintain optimal performance at the lowest cost
    3. ...
    Sources
    [1] Reserved Instances for Amazon EC2 overview 
    [2] What is Amazon EC2 Auto Scaling? 

    To provide targeted recommendations to optimize your EC2 costs, could you tell me:

    1. What's your main goal? (Like Reduce costs for development environments, optimize production workload spending)
    2. Any specific constraints we should know about? (Must maintain high availability, need to keep certain instances running 24/7)

    Guided selection

    When specific valid choices are required, support prompts guide users to select from predefined options.

    Generative AI assistant

    JD
    I need to choose between AWS Fargate and ECS on EC2 for my containerized application.

    Both are great options for running containers. Here are the key differences:

    AWS Fargate:

    1. Serverless - no need to manage infrastructure
    2. Pay only for resources used by containers
    3. Easier to get started and maintain

    ECS on EC2:

    1. More control over underlying infrastructure
    2. Better cost optimization for consistent workloads
    3. Access to instance-level customization
    Sources
    [1] Simplify compute management with AWS Fargate 
    [2] What is Amazon EC2? 

    To provide a more tailored recommendation, what's your primary concern when choosing a container orchestration solution?

    Form-based input

    When multiple specific data points are needed, a dismissible dialog  with various form field types collect additional information.

    Generative AI assistant

    JD
    I need to set up a new EC2 instance

    Here's how to set up a new EC2 instance step by step:

    1. Access the EC2 Dashboard
      • Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/ 
      • Ensure you're in the correct AWS region
    2. Start the Launch Instance Wizard
      • Click on the "Launch instance" button on the EC2 dashboard
    3. Choose an Amazon Machine Image (AMI)
      • Select an AMI that suits your needs (e.g., Amazon Linux 2, Ubuntu, Windows Server)
      • Consider factors like the operating system, pre-installed software, and architecture
    4. ...

    Information required

    I can also help you set up a new EC2 instance. To ensure I configure it correctly for your needs, please provide the following details:

      Linux/UnixWindowsMacOS

      Progressive steps input

      Show follow-up questions in a dismissible dialog  within progressive steps when additional information is needed to complete specific steps.

      Generative AI application

      1. Inspecting resources

      2. Analyzing data

      3. Compiling details

      4. Information required

      Customize your setup

      I can also help you set up a new EC2 instance. To ensure I configure it correctly for your needs, please provide the following details:

      View recent Athena queries 

        General guidelines

        Do

        • To avoid overwhelming the user, minimize the number of follow-up questions in a single workflow, if possible.

        Don't

        • Ask for information already provided or data that can be retrieved automatically through APIs.

        Writing guidelines

        General writing guidelines

        • Use sentence case, but continue to capitalize proper nouns and brand names correctly in context.

        • Use end punctuation, except in headers and buttons. Don’t use exclamation points.

        • Use present-tense verbs and active voice.

        • Don't use please, thank you, ellipsis (...), ampersand (&), e.g., i.e., or etc. in writing.

        • Avoid directional language.

          • For example: use previous not above, use following not below.

        • Use device-independent language.

          • For example: use choose or select not click.

        Component-specific guidelines

        Dialog title

        • Be specific and action-oriented. Dialog titles should clearly indicate the action being performed or the purpose of the dialog, often by combining an action verb with the resource or concept involved.

          • For example: “Configure [specific item]”

        Button text

        • Consistency with title (if applicable). If the button performs the primary action indicated in the dialog title, the verb used in the button label should match the verb in the title.

        • Follow the writing guidelines for button.

        Accessibility guidelines

        General accessibility guidelines

        • Follow the guidelines on alternative text and Accessible Rich Internet Applications (ARIA) regions for each component.

        • Make sure to define ARIA labels aligned with the language context of your application.

        • Don't add unnecessary markup for roles and landmarks. Follow the guidelines for each component.

        • Provide keyboard functionality to all available content in a logical and predictable order. The flow of information should make sense.

        Did this page help you?

        Tell us more - optional

        1000 character(s) available. Do not disclose any personal, commercially sensitive, or confidential information.
        • About
        • Connect
        • Privacy 
        • Site terms 
        • © 2025, Amazon Web Services, Inc. or its affiliates. All rights reserved.
        Made with love atAWS