Skip to content

Latest commit

 

History

History
89 lines (50 loc) · 3.67 KB

features.md

File metadata and controls

89 lines (50 loc) · 3.67 KB

Retail Store Sample Application - Features

The sample application offers several features outside of its functionality a retail store.

Theming

The primary color of the UI can be changed via configuration values, which is useful to provide a distinct appearance.

This is the default theme:

Default theme

This is an alternative theme:

Alternative theme

The theme can be changed in two ways:

  1. Set the RETAIL_UI_THEME environment variable
  2. Append ?theme=<name> to any URL in the UI. This will set a cookie that persists the theme between page reloads.

See the UI component configuration section for valid values for the theme.

Introspection information

This is a page in the UI component that displays information about the environment the container is deployed in. This is useful for demonstrations where it is necessary to show aspects such as cross-AZ load balancing and many others.

The implementation includes metadata for Kubernetes, Amazon EC2, Amazon ECS and AWS Lambda.

This example shows a container running in Amazon ECS:

Metadata page

The page can be found at /info.

To highlight specific attributes you add their names to the URL. For example /info?attr=ecs.cloud.region,ecs.cloud.availability_zone,ecs.cloud.account.id:

Metadata page attributes

You can find the attribute IDs by hovering your cursor over the relevant attribute name in the table:

Metadata page attribute ID

Application topology

This is a page in the UI component that displays information about the application topology thats been deployed. For example it shows what components are deployed, their health and information about their configured dependencies like databases.

Metadata page attributes

The page can be found at /topology.

Generative AI chat bot

This feature provides a chat bot interface directly in the store UI which can be used to demonstrate basic LLM inference use-cases. It is works with Amazon Bedrock and OpenAI compatible endpoints via configuration properties.

Chat bot UI

The chat bot is disabled by default, and can be enabled by setting the environment variable RETAIL_UI_CHAT_ENABLED=true. This will enable the following button to display on the botton-right area of the UI:

Chat bot button

You must also configure one of the below providers.

See the UI component configuration section for more options.

Mock provider

This is designed for quick testing and can be enabled by setting RETAIL_UI_CHAT_PROVIDER=mock. It will return a hard-coded response.

Amazon Bedrock

This provides integration with Amazon Bedrock, the example below shows the relevant configuration values as environment variables:

RETAIL_UI_CHAT_PROVIDER=bedrock # Use Amazon Bedrock
RETAIL_UI_CHAT_MODEL=anthropic.claude-3-5-sonnet-20241022-v2:0 # Provide Bedrock model ID
RETAIL_UI_CHAT_BEDROCK_REGION=us-west-2 # Configure the region to use for Bedrock API calls

OpenAI endpoint

This provides integration with any OpenAI-compatible endpoint, the example below shows the relevant configuration values as environment variables:

RETAIL_UI_CHAT_PROVIDER=openai # Use an OpenAPI compatible endpoint
RETAIL_UI_CHAT_MODEL=meta-llama/Meta-Llama-3.1-8B # Provide a model ID
RETAIL_UI_CHAT_OPENAI_BASE_URL=http://localhost:8000 # (Optional) Base URL for the OpenAI API, usually for self-hosted LLMs
RETAIL_UI_CHAT_OPENAI_API_KEY=<YOUR API KEY> # (Optional) API key for the OpenAPI endpoint